Shoulder Kinematics and Impingement Dynamic Radiostereometric analysis of the shoulder Erling Hallström Academic Dissertation Institute of Clinical Sciences Sahlgrenska Academy University of Gothenburg and the Department of Orthopedics Uddevalla Hospital Gothenburg Sweden 2009 Graphics Graphic form and cover according to institute of Clinical Sciences at Sahlgrenska Academy University of Gothenburg. Cover illustration and anatomy drawings by Catharina Kartus Photographs Ramin Namitabar Drawings Johan Hallström ISBN 978-91-628-7664-7 To Christina, Magnus, Johan and Cecilia CONTENTS CONTENTS 4 LIST OF PAPERS 5 ABSTRACT 6 DEFINITIONS AND ABBREVIATIONS 7 INTRODUCTION 8 Anatomy 9 The kinematics of the shoulder 11 In vitro-studies 11 In vivo – studies 13 Recording methods 15 Impingement syndrome 18 Etiology and pathogenesis 19 Diagnosis 24 Radiographic evaluation 26 Treatment of subacromial shoulder pain 26 Treatment of subacromial impingement 27 AIMS 31 PATIENTS, METHODS AND STATISTICS 32 RESULTS 41 Study I: Shoulder Kinematics in 25 Patients with Impingement and 12 Controls 41 Study II: Kinematic evaluation of the Hawkins and Neer sign 43 Study III: Shoulder Rhythm in Patients and Controls, Dynamic RSA during active and passive abduction. 45 Study IV: Shoulder Kinematics in 19 patients with impingement after arthroscopic surgery, open surgery and physiotherapy. 48 DISCUSSION 50 Repeatability 50 Measurement error-RSA 50 Shoulder joint anatomy and kinematics 50 Kinematic methodology 51 Shoulder motion and age 52 Patient selection 52 Shoulder Kinematics and impingement 52 Neer sign and Hawkins sign 53 The Shoulder Rhythm 54 Effect of treatment 55 The future 56 CONCLUSIONS 57 ACKNOWLEDGEMENTS 59 REFERENCES . 61 APPENDIX 71 PAPERS I-IV 4 List of papers The thesis is based on the following studies, referred to in the text by their Roman numerals I-IV. I Shoulder Kinematics in 25 Patients with Impingement and 12 Controls, E Hallström and J Kärrholm. Clinical Orthopedics and Related Research, 448: 22-27, 2006. II Kinematic evaluation of the Hawkins and Neer sign, E Hallström and J Kärrholm. J Shoulder Elbow Surgery, 17 (15): 40-47, 2008. III Shoulder Rhythm in Patients and Controls, Dynamic RSA during active and passive abduction. E Hallström and J Kärrholm. ACTA Orthop, Accepted for publication IV Shoulder Kinematics in 19 patients with impingement after arthroscopic surgery, open surgery and physiotherapy. E Hallström, T Andrén, A Apelman, A-L Olsson, E Varas, L Virta, Å Öhlund, J Kärrholm. Submitted. COPYRIGHT © 2009 Erling Hallström The copyright of the original papers belongs to the journal or society which has given permission for reprints in this thesis 5 Shoulder Kinematics and Impingement. Dynamic Radiostereometric analysis of the shoulder Erling Hallström Department of Orthopedics, Uddevalla Hospital, Department of Orthopedics, Institute of Clinical Sciences Sahlgrenska Academy at University of Gothenburg Abstract: This study aimed to evaluate the three-dimensional kinematics of the shoulder joint in patients with shoulder impingement and normal volunteers with focus on three well-known diagnostic tests: painful arc test (active abduction), Neer sign (passive elevation) and Hawkins sign. The shoulder rhythm, the speed of motion and whether successful treatment of impingement could be associated with changes of the shoulder kinematics were studied. Dynamic radiostereometry (RSA) is a feasible method for studying rotations and translations of the glenohumeral joint because of its high precision. In all studies the relative motions of the glenohumeral joint was analyzed. In one of them the contribution of the motions in this joint to the absolute or global motions of the shoulder (the shoulder rhythm) was delineated. The median age of the patients and volunteers varied between 49-51 and 30- 36 years in the different studies. 25 patients and 12 subjects without shoulder symptoms were studied during active abduction (painful arc test). The humeral centre displaced medially, proximally, and anteriorly. In the patient group, slightly more (1–1.5 mm) proximal translation was observed in the early phase of the arc of motion. 18 patients and 11 volunteers were tested in the Neer and Hawkins position. In the Hawkins position the centre of the humeral head was positioned more laterally and superiorly in the patients than in the volunteers. In order to analyze the shoulder rhythm and the speed of motion at active abduction, 30 patients and 11 volunteers were studied during active abduction, as well as 21 patients and 9 volunteers during passive abduction, to evaluate the relative and absolute motion. The patient group showed more scapular and trunk motions (p=0.04) and especially up to 40°. The distribution of motion between the glenohumeral joint and the trunk in both patients with impingement and volunteers was less than or equal to 1:1. 19 patients were randomized to three treatment options: physiotherapy (n=7), open surgery (n=7) or arthroscopic surgery (n=5). RSA studies and clinical evaluation were done before and median 29 and 24 months later. According to Constant-75, patients treated with surgery improved significantly more than those treated with physiotherapy (p<0.05). In the total material there was a tendency to increasing Constant score with increasing medial and posterior position of the humeral head center in a test for Hawkins sign. In conclusion, the patients showed an increased proximal translation in the painful arc test and when placed in the Hawkins position a more lateral and posterior position of the humeral head center. The glenohumeral-thoracoscapular ratio was less than or equal to 1:1 in patients and volunteers, where the patients had reduced glenohumeral motions in the early phase of active abduction. Correlation between changed humeral head translation after treatment during the test for the Hawkins sign and improvement of the Constant-75 score in the total patient material might represent a causal relationship, but these findings need to be further studied in larger patient groups. Keywords: Shoulder kinematics, radiostereometry, impingement, open surgery, arthroscopic surgery, physiotherapy, clinical outcome 6 Definitions and abbreviations AP anterior-posterior Abd abduction Add adduction IR internal Rotation ER external Rotation RSA RadioStereometric Analysis SEM Standard Error of Mean UmRSA RSA software provided by RSA Biomedical, Umeå, Sweden OpenMRI open Magnetic Resonance Imaging 7 Introduction and background The term “impingement syndrome” was coined by Neer in 19721. He came to the conclusion after studies in the anatomy laboratory and at surgery that impingement occurs against the anterior edge and undersurface of the anterior third of the acromion, the coracoacromial ligament, and on occasion, to the acromio-clavicular joint. Physical signs and symptoms include the impingement sign, arc of pain, crepitus and varying weakness. Neer2 proposed three stages in the progress of impingement syndrome: stage I, a reversible stage in which edema and hemorrhage dominate; stage II, an irreversible stage in which tendinitis and fibrosis have occurred; stage III characterized by tendon degeneration and tearing. In all three stages of impingement the symptoms are almost identical but in stage III with advanced cuff rupture the weakness of the shoulder is more pronounced. Impingement syndrome of the shoulder is by many believed to be the most common cause of shoulder pain3 and accounting for half of the patients consulting physician because of shoulder pain4-6. Impingement syndrome is thought to be caused by inadequate space for clearance of the rotator cuff tendon as the arm is elevated7-9. Kinematic changes are believed to occur primarily in symptomatic patients and to result in additional decrease of the subacromial space, which could aggravate the symptoms7, 10, 11. Motions that bring the greater tuberosity closer to the coracoacromial arch may be particularly problematic. These motions include excessive superior or anterior translation of the humeral head on the glenoid fossa, inadequate lateral (external) rotation of the humerus, and decreased normal scapular upward rotation12-16. These theories were questioned by Budoff et al17 who thought that 90%-95% of all rotator cuff abnormalities could be attributed to intrinsic breakdown of the rotator cuff tendon because of tension overload, overuse, and traumatic injury rather than mechanical compression. Other factors such as ischemia and degeneration related to age and overload of the short rotator muscles may contribute to the complaint18. Although controversial, most authors acknowledge that compression is one of the factors, which can lead to rotator cuff pathology19, 20. Physiotherapy with a physical training program is suggested as an initial treatment option for patients with impingement syndrome8, 21-23. Böhmer24 described a new physiotherapy method in 198424 which has gained interest and been recently described in a study by Virta et al25. One of the tools used by Böhmer24 was a Sling. Järvholm26 demonstrated that the load on the supraspinatus muscle was reduced by approximately 30% by the use of a sling. Correct instruction and feedback given by the physiotherapist in order to correct dysfunction in the shoulder rhythm and motivate for regular exercise may be crucial in the method described by Böhmer24. Experimental and observational studies have described that the subacromial space is influenced by muscle activity. A study14 examined the subacromial space by MRI of four normal volunteers and found that it was narrowed by protraction and widened by retraction of the scapula. In a cadaver study Wuelker et al27 observed that lack of force in the infraspinatus, teres minor and subscapularis increased the subacromial pressure by 61%, whereas lack of force in the supraspinatus muscle did not significantly alter the subacromial pressure. It is commonly described that the purpose of the surgical treatment is to enlarge the subacromial space and thus decompress the subacromial structures28. This study aimed to evaluate the three-dimensional kinematics of the shoulder joint in patients with shoulder impingement as compared to volunteers with a focus on three well- 8 known diagnostic tests: painful arc test (active abduction), Neer sign (passive elevation) and Hawkins sign (passive abduction, flexion and internal rotation). The shoulder rhythm, the speed of motion and whether successful treatment of impingement could be associated with changes of the shoulder kinematics were studied. Anatomy The shoulder joint has greater range of motion than any other joint in the body29. For normal function a complex interaction between the muscles and four articulations, the stenoclavicular, acromioclavicular, glenohumeral and scapulothoracic joint is required (Figure 1). Figure 1. The four articulations of the shoulder, the sternoclavicular, acromioclavicular, glenohumeral and scapulothoracic joint. The articular bony surface of the glenoid measures only 33% of the humerus30. It is less curved than the surface of the humerus29, 31, 32. Consequently, the glenohumeral joint can function not only as a ball-in-socket joint, but in a more compound mode. The humeral head cannot only rotate but also translate33-35. The articular surface of the glenoid and the labrum constitute an opening that is about 9 mm deep in the proximal/distal direction and 5 mm deep in the anterior-posterior direction. According to Saha36 the labrum widens the glenoid surface to embrace 75% of the humeral head vertically and 57% horizontally. Because of the small bony surface of the glenoid the stability of the shoulder is to a substantial extent dependent on the ligaments and surrounding muscles. At zero degrees of abduction of the humerus, the main stabilizer of the glenohumeral joint to counteract inferior humeral subluxation and anterior/posterior tension is the superior glenohumeral ligament (SGHL37, Figure 2). When the arm is in less than 90° of abduction, the middle glenohumeral ligaments (MGHL) reduces the external rotation, but otherwise it does not have an effect on the movement of the arm. 9 The inferior glenohumeral ligament (IGHL) is, according to O´Brien et al37, arranged in three segments: an anterior thick part, a posterior somewhat thinner but well defined part, and a still thinner overriding axillary purse creating a “hammock-type” model. With external rotation, “the hammock” glides anteriorly and superiorly. The anterior part is stretched and the posterior part relaxes. The reverse occurs at internal rotation. When the shoulder is abducted to 45° the anterior (IGHLa) and inferior (IGHLb) glenohumeral ligament is the main stabilizer to anterior and posterior tension (Figure 2). From the coracoids process the coracohumeral ligament (LCH) expands nearby the plane of the capsule to the tuberosity between the supraspinatus and subscapularis tendons and continues to the tendinous insertion of the cuff. The coracoacromial ligament (CAL) extends from coracoids to the anterior acromion. Together with coracoids, the anterior acromion and the distal part of the clavicle it creates the coracoacromial arc under which the supraspinatus extends to reach its insertion at the greater tuberosity1, 2. The supraspinatus (SSP) tendon and the tendons of the subscapularis (SCL), infraspinatus (ISP), and teres minor (TM, Figure 3) insert into the underlying glenohumeral capsule near the greater tuberosity1, 2. The space between the humeral head and the anterior- inferior edge of the coracoacromial arch has been named the supraspinatus outlet by Neer and Poppen38. Figure 2. The ligaments of the glenohumeral joint. The muscles of the shoulder work as a dynamic stabilizer where the rhomboids (RH), levator scapulae (LSC), trapezius (TR) and serratus anterior (SERA), which also control movements of the scapula, interact with the muscles of the rotator cuff (subscapularis, supraspinatus, infraspinatus and teres minor). The deltoid (DLT), pectoralis major (PMA), latissimus dorsi (LAT), and the biceps brachi control (BB) the glenohumeral joint in a synergistic way to achieve joint compression (Figure 3)33, 34, 39-44. 10 Figure 3. The muscles of the glenohumeral joint. The Kinematics of the shoulders In vitro studies Harryman et al33 studied 8 shoulders with disarticulation of the scapulothoracic and sternoclavicular joints, using of a load-cell and a spatial sensor with six degrees of freedom. The receiving coil of this sensor was rigidly attached to the humeral shaft as close to the humeral head as possible without interfering with a full range of motion. With flexion of the glenohumeral joint they recorded anterior translation - and with extension posterior translation - of the humeral head. Posterior translation was also observed during external rotation. During adduction (cross-body movement) the sensor at the humerus shifted anteriorly. Despite application of a posteriorly directed force of 30 to 40 Newton the anterior translations occurred with flexion and consequently could not be avoided. In one experiment the posterior portion of the capsule was tightened by suturing. After this procedure anterior translation was increased during flexion as well as during adduction (cross-body movement). The translations also started earlier in the arc of motion. At the same time a more proximal translation during flexion of the glenohumeral joint was registered. Thompson et al45 used 8 fresh-frozen full upper extremities acquired from human cadavers to evaluate the effects of rotator cuff deficiency on shoulder biomechanics and humeral translations in a cadaveric model. The dynamic shoulder testing device consisted of 6 11 servo-actuated hydraulic cylinders to apply forces to each of the rotator cuffs and the middle deltoid tendons through a tendon clamp-cable-pulley system. To accurately represent the extremity mass distribution an intact human upper extremity was used. Glenohumeral joint motion was measured using a six degree-of-freedom magnetic tracking device with an accuracy of 0.8 mm and 0.8° 46. No statistically significant difference existed between muscle force for the 1, 3 or 5 cm full thickness rotator cuff tears, provided that the subscapularis, infraspinatus and teres minor tendon remained intact with full glenohumeral abduction. The translation of the humerus with respect to the glenoid was found to be less than 2.0 mm in all three planes (anterior-posterior, medial-lateral, proximal-distal). Also Wuelker et al47 used 8 cadaveric samples to study the translation of the centre of the humeral head with simulated active elevation. The deltoid muscle and the rotator cuff were connected to a controlled hydrodynamic actuator through wire cables. The glenohumeral joint was elevated 90° using a constant force. An ultrasonic device was used to determine the position of the arm in all spatial directions. The author found mean translations of the centre of the humeral head to be 9.0 mm ± 5.2 superiorly and 4.4 mm ±1.3 anteriorly between 20 and 90° at active elevation of the joint. Later the Wuelker et al27 used an electronic device (capacitive sensor) to evaluate the subacromial pressure with use of the same set-up. The peak pressure (mean 57 N/cm³) was in the majority of the samples recorded at the anterior border of the acromion. Relaxation of the supraspinatus muscle showed an 8% decrease of the mean coracoacromial pressure, whereas relaxation of the subscapularis and infraspinatus/teres minor and the rotator cuff resulted in a corresponding increase of 61% and 35%, respectively. After anterior acromioplasty the mean coracoacromial pressures decreased by only 5%. Zuckerman et al9 measured a subacromial width of 6-7 mm, but under the coracoacromial ligament it was only 1.5 mm. 10 years later Maeskers et al48 described three-dimensional geometry of the supraspinatus outlet in 32 cadaver samples. They measured the maximum mean width to be about 6.0 mm and found a range between about 1.4 to 6.9 mm. They also observed how the size of this outlet changed with changes to the relative orientation of the humerus with respect to the scapula during motion of the arm in the frontal and sagittal plane in 10 normal volunteers49. The geometrical and kinematic data were combined to study the supraspinatus outlet during elevation of the humerus in the frontal and sagittal plane. Throughout arm elevation, the greater tuberosity was shifted away from the coracoacromial arch resulting in narrowing of the outlet during elevation in the frontal plane but only from 60° to 120°. The variations between sequential trials and individuals were large, caused by differences in anatomy and pattern of motion. Both Zuckerman et al 9 and Maeskers et al 48 noted that the critical zone of the supraspinatus outlet is located under the coracoacromial ligament and not under the acromion. Billuart et al50 studied the kinematics of the glenohumeral joint in 6 cadavers using an optoelectronic system. The humerus was moved by pull of the deltoid without constraining the humerus. By pulling horizontally in the anterior and medium fibers of the deltoid the glenohumeral joint was abducted 24° to 30.5°, minimally flexed (1.5°) or extended (31°), externally (12°) or slightly internally rotated (5°). The humeral head translation along the three coordinate axes was less than 5 mm. They concluded that these results propose that the deltoid alone can abduct in the glenohumeral joint with maintained stability of the joint. 12 In vivo studies The complex interplay between the scapula and the glenohumeral joint during elevation of the arm was probably first described in 1884 by Cathcart (quoted from Crosbie et al51), but the term “scapulohumeral rhythm” which describes this phenomenon appeared later52. Inman, Saunders and Abbot 194441 studied living subjects both with radiography and after insertion of pins into the bone during active abduction. They concluded that this motion involved four joints: the sternoclavicular, the acromioclavicular, the scapulothoracic and the glenohumeral joints, that in a synchronous way acted together. They found that in between 30° to 60° of elevation the scapula and humerus try to find a position of stability which they may obtain in one of numerous ways. Motion may take place in the glenohumeral joint while the scapula stays fixed until stability is reached. Alternatively, the scapula may translate laterally or medially on the chest wall, or it may move back and forth to find a stable position. This early phase of motion was labeled the “setting action” or “setting phase”. Once 30° of abduction or 60° of flexion had been accomplished, the relation between scapular and humeral motion remained constant. Thus, with further motion, they found the ratio of one to two corresponding to 10° of glenohumeral motion for every 15° of motion of the arm. Overall the entire contribution of scapular motion to the total amount of elevation never surpassed 60°. Doody et al53 evaluated the contribution of scapular motion to shoulder abduction a 30° intervals in 25 women with goniometric technique. Abduction in the scapular plane with and without resistance was studied. With no added stress the mean scapular and glenohumeral contributions were 59° and 113°. Freedman et al54 evaluated the scapular and glenohumeral movements in the scapular plane during abduction of the arm. Fifty-two male medical students were studied with their arms in five positions (0°, 45°, 90°, 135° and at maximum elevation) in the scapular plane corresponding to an angle of 30° in relation to the coronal plane. The ratios of glenohumeral to total arm movement (GH/A) and glenohumeral to scapular movement (GH/S) for the four intervals between the five positions were calculated (Table 1). The ratios were rather similar up to 135° of arm abduction. Past 135° the ratios increased indicating increasing motion in the glenohumeral joint Interval GH/A GH/S 0-45° 0.589 1.431 45-90° 0.58 1.379 90-135° 0.556 1.253 135-max 0.738 2.729 Table A. Ratio of glenohumeral/arm (GH/A) and glenohumeral/scapula (GH/S) motion during abduction according to Freedman et al54. Poppen at al12 used radiography to evaluate the movement of the arm and the relationship between the scapula and the glenohumeral joint motion during abduction in the plane of scapula (elevation). The centre of rotation of the glenohumeral joint was also studied. 12 volunteers without shoulder symptoms and 15 patients took part in the study. Between 0°- 30° most of the motion was in the glenohumeral joint. After about 30° of abduction they recorded a ratio of glenohumeral to scapulothoracic movement corresponding to 5:4. The centre of rotation of the glenohumeral joint during elevation was located within a 6-mm distance from the geometric centre of the humeral head. For every 30° of elevation the average proximal/distal translation of the humeral head was less than 1.5 mm in the volunteers. 13 Prior injury to the rotator cuff resulting in impaired function of the shoulder joint was associated with abnormal translation of the instant centre of rotation. The instant centre of the humeral head showed from 0 to 30°, and often from 30° to 60° this centre displaced about 3 mm proximally in the patients. Thereafter it shifted 1-2 mm proximally or distally between the sequential positions studied. In the volunteers the average movement from position to position was only about 1 mm. Howell et al34 studied the relationship of the humeral head to the scapula in the horizontal plane of motion on axillary roentgenograms in volunteers and patients with anterior instability were evaluated. In the control group the humeral head was center in the glenoid throughout the horizontal plane of motion except when the arm was in maximum extension and external rotation then the center of the humeral head was rested approximately 4 mm posterior to the centre of the glenoid cavity. When the arm was flexed or rotated from this the humeral head displaced anteriorly. Paletta et al55 evaluated glenohumeral kinematics and glenohumeral-scapulothoracic motion with two-plane radiography in patients with anterior instability or rotator cuff tear before and after surgical treatment and rehabilitation. 6 normal adults constituted a control group. 18 patients with anterior shoulder instability (group A) and 15 patients with rotator cuff tears (group B) were studied before surgery. In 7 of 18 patients with anterior instability and in all with rotator cuff tears there was demonstrated a superior translation during scapular plane abduction. Anterior translation of the humeral head was only found in cases with instability (14 of 18). Both groups of patients demonstrated an altered relationship between glenohumeral and scapulothoracic motion compared with the control group. In group A all patients studied (n=12) and 12 of 14 in group B demonstrated normal glenohumeral kinematics in both planes after open anterior stabilization or rotator cuff repair. In group A the changed relationship in the glenohumeral-scapulothoracic motion persisted, whereas in group B this relation became normal. Grachien et al56 studied humeral head translation during passive and active elevation using an open MR technique and 3D digital post processing methods. 15 normal volunteers were examined with an open MR system at different abduction positions under muscular relaxation (30°-150° of abduction) and during loading of the shoulder muscle (1kg, 60° - 120°). Their relative positions were calculated after segmentation and 3D reconstruction and the centre of the glenoid and the midpoint of the humeral head were analyzed. During passive elevation, the humeral head translated about 1 mm inferiorly and 1.5 mm posteriorly from 30° to 150°. During loading of the shoulder muscle the humeral head obtained a more inferior position and was more centered, particularly at 90° and 120° of abduction. Along the AP axis the humeral head was more centered at 60 and 90° of abduction during loading of the shoulder. The authors concluded that neuromuscular control was of important to achieve stability of the joint. Later these authors57 studied glenohumeral-scapulothoracic motion and the supraspinatus muscle with the same technique. 14 volunteers were examined in 5 positions of abduction (30°-150°) The axis of the supraspinatus, humerus, clavicle, and the plane of the glenoid were determined, and the relative movements were calculated. The ratio for glenohumeral to scapulothoracic motion was 1.5:1 at 60° and 2.4:1 at 120° of abduction. At 30° the axis of the supraspinatus was nearly horizontal but tilted with further abduction to reach slightly more than 120° at 150° of abduction. In the transverse plane, the angle between the supraspinatus and the clavicle axis became larger during abduction because of an increasing retroversion of the clavicle. In further studies Grachien et al58 evaluated 20 patients with unilateral impingement and 14 volunteers without shoulder symptoms at 30°, 60° and 120° of abduction with and without abducting muscle activity. There were no major differences in glenoid rotation between the 14 patients and the healthy subjects. Comparison between muscle activity and muscular relaxation showed no major difference between the groups. The ratio of glenohumeral to scapulothoracic motion (relative to the spinal axis) showed no large differences between the control group and the affected shoulders of the patients. The authors concluded that patients with various stages of impingement syndrome did not display altered motions on the affected side under the given conditions of these examinations. These findings applied both to elevation with and without muscle activity. They also reported a scapulothoracic-glenohumeral motion ratio between 1:1.8 and 1:2.4, which corresponded to previous studies12, 54, 55. Finally this group59 analyzed the effect of abducting and adducting muscle activity on glenohumeral translation, scapular kinematics and subacromial space in healthy volunteers at 30°, 60°, 90°, 120° and 150° of arm elevation. A force of 15 N was applied to the distal humerus, to obtain isometric contraction of the abductor and adductor muscles. Adducting muscle activity resulted in an increase of the subacromial space in all arm positions, whereas the scapular-humeral rhythm (2.2-2.5) and scapular tilting (2-4°) remained relatively constant during elevation without any large variation between abducting and adducting muscle activity. Comparison between adduction and abduction in midrange elevation (60-120°) revealed that the position of the humerus in the former position was more inferior and anterior. Thus, the subacromial space could be widened by adducting muscle activity. Mesckers et al48 also observed how the size of this outlet changed in 10 normal volunteers with changes in the relative orientation of the humerus with respect to the scapula during motion of the arm in the frontal and sagittal plane. Throughout arm elevation, the greater tuberosity was shifted away from the coracoacromial arch resulting in a narrowing of the outlet during elevation in the frontal plane but only from 60° to 120°. The variations between sequential trials and individuals were largely caused by differences in anatomy and pattern of motion Ebaugh et al60 evaluated the effects of active and passive arm elevation on the scapulothoracic motion in 20 subjects without any history of shoulder problems. The motion was calculated from electromagnetic sensors adapted to the scapula, thorax and humerus in three dimensions. Muscle activity from the upper and lower trapezius, serratus anterior and posterior deltoid and infraspinatus was recorded with surface electrodes. They found a more upward and external rotation of the scapula, clavicular retraction and elevation, especially during active compared to passive elevation between 90° and 120°. They concluded that especially throughout the mid-range of arm elevation, the upper and lower trapezius and serratus anterior muscles seemed to have an important role for scapular rotation. Recording methods of shoulder kinematics Since 1899, physical models of the shoulder have been used in attempts to replicate glenohumeral joint motion and to study the function of tendons and muscles62-64. These early models replaced muscles with cords whose change in length during motion of the bones at the shoulder were quantified. Later work examined the contribution of the static restraints (osteoarticular surface, capsulo-ligamentous structures, and weight of bone and soft tissue) to achieve glenohumeral joint stability33, 65, 66. Commonly used methods to measure joint motion such as film and video recordings of markers glued to the skin are less accurate, because most of the scapula is surrounded by a comparatively thick soft tissue envelope. Nonetheless, such methods have also been used also rather recently. The methods used in recording the shoulder kinematics can be divided into the following subgroups, based on the different procedures and tools used when carrying out the 15 investigation. The subgroups are skin-based methods, methods evaluating the shadows, pins in the skeletal area, use of external devices measuring the kinematics (goniometry), conventional radiography, magnetic resonance imaging (MRI) and radiostereometric analysis (RSA). Markers/transmitters fixed to the skin The Electromagnetic sensor device has been used to measure the three-dimensional shoulder movement and has been described by Johnson et al67. This is a method based on spherical polar coordinates previously described by Kapandji68. The 3-space Isotrack (Polhemus navigation Systems, U.S.A.) is an electromagnetic sensing device for the measurement of the location and orientation of a sensor in space, each connected to an electronic unit including the hardware and primary software for data collecting control. The source (transmitter) generates an electromagnetic field (detected by the sensor) and the electronics package computes the relative location and direction from the detected magnetic field with the full 6 degrees of freedom. The system is associated to a PC which calculates the results obtained by specifically designed software. The sensors are attached to the skin and can be mounted at the sternum, scapula and humerus using adhesive tape. Subjects are standing anterior to the transmitter. With a high speed reflex camera (Bolex model H16 high speed 16-mm reflex camera) Bagg et al69 was evaluated the scapular rotation during arm abduction in the scapular plane. Both arms were moved in the plane of the scapula by placing the subject`s forearms and palms against two vertical guiders that were aligned 30° anterior to the coronal plane. Specific landmarks were (1) the root of the scapular spine, (2) the acromial angle and (3) the inferior angle of the scapula. Two markers were also placed along the long axis of the humerus. In addition, reference markers were positioned over the spinous process of several vertebrae. During film analysis, the three markers of the scapula, the two on the humerus and two of the lower vertebral reference markers were digitized on a Vanguard model M-16C motion analyzer. Programs were then developed to measure the scapular humeral angles and to estimate the location of the scapular ICR for each 15° increment of arm abduction. Shadow-based methods Moiré topography (Figure 4) is a form of biostereometry, and has been very useful in describing the three-dimensional character of the human body70. The subject is positioned behind a grid of horizontal lines which is illuminated by a light source. An optical effect is seen when the line shadow through the grind conforms to the surface topography of the subject. Edging models are formed that appear as contour lines on the subject. The contour lines of the surface will accurately reflect the asymmetry of the scapulothoracic area as long as the subject`s back is kept parallel to the apparatus70. Figure 4. Moiré topography. 16 During testing the subjects are positioned with their backs approximately 1 cm from the apparatus. Then static and dynamic testing of the shoulder is possible to achieve in a proper way. For minimizing technique variation, one tester supervised all Moiré evaluations and photographs were taken by one photographer. All Moiré photographs were evaluated according to a uniform measurement technique70. Radiography Radiographic studies in which the positions of the scapula and clavicle have been projected on roentgen films at various angles of humeral abduction or anteflexion12, 54, 180 Magnetic resonance imaging Recently several studies have been done with an open MRI 58, 59, 71-74. The subject is placed in a supine position where the investigator is able to study the shoulder girdle in various positions but dynamic procedure is not possible to obtain. Images are acquired by use of a 3D gradient recalled echo (GRE) pulse sequence with 20 milliseconds of echo time, 37 milliseconds of repetition time, a 20 x 20 cm field of view, and a 256 x 160 – pixel matrix. Each 3D GRE scan yielded 42 consecutive 2 dimensional images with a segment deepness of 2 mm and needed a total scan time of 4 minutes 34 seconds74. Radiostereometric analysis Radiostereometric analysis (RSA) is an alternative approach in studying kinematics61, 163- 4.This method is based on fixed skeletal landmarks and has a documented high resolution. It has been frequently used in evaluating the migration and wear of prosthetic implants165, 166. Dynamic radiostereometry75, 76 also enables recordings during active joint motion. This application has gained interest in the last decades167, 168, 170-1. The technique has also been used in evaluating knee motions in different aspects after total arthroplasty169. Under local anesthesia 4-6 spherical tantalum markers (Ø = 0.8 or 1.0 mm) were inserted into the scapula (acromion) and the humeral head. A set up including two film-exchangers, placed side by side designed for simultaneous exposures (Figure 5). The exposure rate was set at 2 per second during 5 seconds during the abduction/elevation of the arm. The radiographic examination was initiated with a starting or reference position corresponding to a well defined anatomic position. All subsequent recordings were related to this position of the arm. A fictive point corresponding to the humeral head centre was constructed by circular templates to enable measurements of humeral head translations in a reproducible way. The X- ray films were scanned at 300 dpi using a flat-bed scanner (Sharp JX610, Osaka, Japan) and measured using dedicated software75. The data were then analyzed by specifically designed software (UmRsa). Other methods Goniometers and pins inserted into clavicle and scapula and other bones have been used to measure externally the motion of the bones used53, 193-4. 17 Figure 5. A set up of X-ray tubes and two film-exchangers, placed side by side designed for simultaneous exposures. Impingement syndrome Epidemiology Shoulder pain is a common disorder. In the general population the prevalence of shoulder pain may be as high as 6 - 11% under the age of 50 years, increasing to 16 - 25% in the elderly77-79. Estimates of the annual incidence of shoulder disorder incurred in general practice varies from 7 - 25 per 1000 registered patients per year5. Inability to work and to carry out household activities in addition to loss of productivity can become a considerable burden to the patient as well as to society80. Shoulder impingement syndrome In the past, many authors described abnormal conditions in the subacromial space1, 2, 52, 85, 86, but the true reason why impingement syndrome develops remains unclear. In 1931 Meyer87 proposed that because of the friction between the rotator cuff and the undersurface of the acromion, tears of the rotator cuff could develop secondary to attrition. He also described tears close to the greater tuberosity, but did not explain their etiology. In 1934 Codman52 focused on a specific and vulnerable location on the rotator cuff, situated one centimeter medial to the insertion of the supraspinatus on the greater tuberosity where most of the degenerative changes were found. Neer1 described shoulder impingement as a mechanical phenomenon corresponding to impingement of the rotator cuff tendon beneath the anterior-inferior acromion. This condition occurs when the shoulder is placed in forward flexion and internal rotation. He hypothesized that the rotator cuff is impinged by the anterior one-third of the acromion, the coracoacromial ligament, and the acromioclavicular joint. Neer also proposed that the insertion of the supraspinatus tendon on the greater tuberosity is involved in the impingement conditions. In addition he suggested that these tears also could be caused by bony spurs in the coracoacromial ligament. 18 In 1983 Neer2 characterized 3 stages of impingement. Stage I is described as edema and hemorrhage of the bursa and the rotator cuff, a common disorder among patients who are less than 25 years old. Stage II represents permanent changes such as fibrosis and tendinitis of the rotator cuff, and is normally found in patients who are 25–40 years old. Stage III corresponds to more chronic changes, such as partial or complete tears of the rotator cuff. It is usually seen in patients who are more than 40 years old. Although the more advanced stages of this process, including rotator cuff tears are more common in older individuals, impingement and rotator cuff pathology are also frequently seen in younger, athletic individuals, who are engaged in repetitive overhead activities or in young workers, who expose their rotator cuff to similar conditions. Etiology and pathogenesis Shoulder impingement can be divided into external and internal categories. External impingement is caused by structural changes outside the joint and includes primary, secondary and subcoracoid types. Internal impingement is secondary to rotator cuff and capsular dysfunction. It may be divided into 4 types: posterior-superior (classic internal impingement), anterior-superior, anterior and entrapment of the long head of the biceps tendon. Extra articular impingement (External impingement) It is important to be familiar with the anatomical features of the subacromial space to understand the pathogenesis of subacromial impingement. The superior limit consists of the coracoacromial arch: the acromion, the coracoacromial ligament and the coracoids process. The acromioclavicular joint is directly superior and posterior to the coracoacromial ligament. The inferior limit consists of the greater tuberosity of the humerus and the superior part of the humeral head. According to radiographic measurements, the height of the space between the acromion and the humeral head varies between 1.0 – 1.5 cm81. The rotator cuff tendons, the long head of the biceps tendon and the bursa are localized in the subacromial space. This reduces the subacromial space considerably more than seen on radiographs. Impingement may develop due to any deviation that changes the relationship of these subacromial structures. Primary impingement is considered to be caused by degenerative changes of the acromioclavicular joint or due to certain variations of the acromial morphology. Secondary impingement may develop due to elevation of the humeral head and/or joint laxity and instability. Impingement occurs when the supraspinatus tendon is squeezed in the supraspinatus outlet space. It starts as an inflammatory process involving the subacromial bursa and the tendon itself. Bursitis, inflammation, edema and hemorrhage may develop. If the process continues, fibrosis of the subacromial bursa and tendinitis may develop. Further progression of this condition may evolve to partial or full thickness tears of the supraspinatus tendon1. Acromial morphology in shoulder impingement syndrome Differences in the shape and slope of the acromion were described as early as 190988. In the past the treatment of the shoulder pain caused by subacromial impingement was mainly focused on removing different fractions of the acromion. Armstrong85 suggested that total acromionectomy would relieve these symptoms as did Diamond81. McLaughlin and Asherman89 proposed that lateral acromionectomy would be sufficient. These treatment options were not the solution and numerous complications ensued, especially detachment of the deltoid muscle. 19 In 19721 Neer proposed that differences in the shape and the slope of the anterior portion of the acromion could explain subacromial impingement and associated tears of the rotator cuff. These conclusions were based on his own clinical remarks as well as dissection of more than 100 cadaveric scapulae. In addition, a spur on the coracoacromial ligament was often found distally directed into the subacromial area. Bigliani et al90 described three frequently observed variations of the morphology of the acromion based on cadaveric dissections and radiographs. In 139 shoulders from 71 cadavers they identified 3 types of morphology. Twenty-four (17%) were rather flat, 60 (43%) were described as curved, and 55 (40%) as hooked. A higher prevalence of full thickness tears of the rotator cuff was noted in association with the hooked or type III acromion. This observation was confirmed by Morrison et al 198791, who studied 200 consecutive patients with supra outlet radiographs. Sixty-six (80%) of the 82 patients who had rotator cuff tear according to arthrography had a hooked acromion. Morrison et al92 confirmed these observations and preferred supraspinatus outlet radiographs to MRI to arrive at correct diagnosis. In 1986 Aoki93 et al described that presence of spurs was common with the flat type of acromion and these cases had increased pitting on the surface of the greater tuberosity. In the 130 specimens studied they found that the prevalence of spurs in the subacromial space increased with age. Nicholsson et al94 studied 420 specimens and noted that the prevalence of spur formation at the anterior part of the acromion increased after 50 years of age, whereas the morphology of the acromion did not seem to change with age. The hypothesis that the anterior part of the acromion is associated with the pathogenesis of tears in the rotator cuff was supported by Zuckerman et al9. They studied 140 cadaveric shoulders and found that the supraspinatus outlet was 22.5% smaller and the anterior projection of the acromion was larger in the specimens with rotator cuff tear. Rockwood and Lyons95 emphasized the importance of the anterior prominence of the acromion in impingement syndrome. They suggested a two step acromionectomy, resection of the anterior part of the acromion at the level of the clavicle and removal of bone from the inferior aspect of the acromion. Despite these studies the association between the morphology of the acromion and supraspinatus pathology has been questioned, mainly as it is related to the reliability of those classifications presented. Poor levels of interobserver reliability96 or a more complex and subtle variation of the acromial shape in association with difficulties to obtain a representative image of its true shape on MRI or radiographs has been debated. More recently, a study by Chang et al97 used complex 3D computer modeling of the acromial undersurface. He concluded that the shoulder impingement or rotator cuff tears are not primarily caused by osseous impingement by the acromion. Impingement by the Coracoacromial ligament The “snapping shoulder,” a condition starting with shoulder pain is believed to be caused by inflammation and swelling of the subacromial bursa, which becomes squeezed under the edge of the coracoacromial ligament, was described by McLaughlin and Asherman89. Later, Neer 1, 2 incorporated resection of the coracoacromial ligament as an essential part of the anterior acromioplasty procedure. This procedure has also been recommended by others and especially in athletes engaged in overhead activities98-101. In a cadaveric study, Burns and Whipple102 noted that the supraspinatus and biceps tendons were stabbed against the lateral edge of the coracoacromial ligament as the arm was flexed forward to 90° and then forcibly internally rotated. Soslowsky103 proposed that enlargement of the coracoacromial ligament could result in subacromial impingement. However, this 20 hypothesis is questioned by Sarkar104 and Uthoff105, who in histological studies only found degenerative changes without any swelling of this ligament. Degeneration of the Acromioclavicular (A-C) joint Degenerative changes of the A-C joint are a widely accepted reason for subacromial impingement1, 2, 106-108. When the cuff passes underneath the joint, osteophytes from the lateral end of the clavicle or from the medial part of the acromion in the A-C joint extend beyond the A-C joint and interfere with the rotator cuff. Kessel and Watson106 found that the pain disappeared in about 2/3 of 97 patients with “painful arch” syndrome after local injection of anesthetics and a steroid or after division of the coracoacromial ligament. These patients had lacerations of either the anterior or posterior part of the rotator cuff. In the remaining patients with degenerative changes in the A-C joint, excision of the distal part (1 cm) of the clavicle resulted in pain relief for patients. Osteoarthritis of the A-C joint may be one reason for unsuccessful operative treatment of subacromial impingement. However, resection of the lateral clavicle should only be done if the patient has symptoms localized to the AC-joint in combination with radiographic changes in this region20. Subcoracoid impingement In subcoracoid impingement the subscapularis tendon, the subcoracoid bursa, and the anterior joint capsule is squeezed between the coracoids and the lesser tuberosity. In this condition the distance between the coracoid and the lesser tuberosity (the coraco-humeral interval) is considered to be narrowed due to lengthened coracoids. This can be a hereditary condition or post-traumatic with deformity of the coracoid or the humeral head or iatrogenic as a consequence of glenoid osteotomy or coracoplasty109. The normal coraco-humeral interval has been shown to vary between 8 and 11 mm84, 110, 111. It is smaller in females than in males84, 112. Subcoracoid stenosis has been defined as a coraco-humeral interval less than 6 mm. As measured on CT scans the coraco-humeral interval decreased from about 9 to 7 mm when the arm was placed in flexion and internal rotation, a position recognized to induce subcoracoid impingement84. When the coraco-humeral interval was evaluated by MRI (axial view) the mean distance amounted to about 10 mm. In a group of patients with torn subscapularis tendons this distance decreased to about 5 mm194. A “roller-wringer effect” has been thought to cause tears in the subscapularis tendon in patients with subcoracoid impingement110. The coracoid displaces the surface of the tendon during rotation of the shoulder and performs a roller-like action, which induces progressive damage, eventually leading to macroscopically visible tears. Giaroli et al112 tried to evaluate if the distance of the coraco-humeral interval as measured on routine shoulder MRI could be used as a diagnostic tool for this condition, but came to the conclusion that subcoracoid impingement was primarily a clinical diagnosis, which can only may be confirmed by MRI. Os Acromiale In 1863 Gruber described os acromiale, a bony formation corresponding to a remaining separated distal acromial epiphysis (Quoted from Bigliani 1997)20. The prevalence of this condition has been estimated within a range of 1 - 15%113, 114. Axillary radiographs may be necessary to observe this bone. An os acromiale might be movable. This bone may also slope anteriorly and cause impingement115, 116. 21 Intra-articular shoulder impingement (Internal impingement) Internal impingement syndrome involves the intra-articular surface fibers rather than the bursal surface fibers of the rotator cuff. Posterior-superior internal impingement was originally described by Walch, who observed undersurface tears of the supraspinatus and infraspinatus tendons between the posterior-superior glenoid rim and the humeral head117. Internal impingement also includes anterior-superior impingement, anterior impingement, and entrapment of the long head of the biceps tendon. Posterior-superior impingement syndrome The posterior-superior impingement syndrome is defined as a condition where the intraarticular side of the supraspinatus and infraspinatus tendons is impinged on the posterior edge of the glenoid when the arm is in abduction and external rotation117. It causes rotator cuff undersurface tears in athletes during the late elevating-early speeding up phase of overhead movement117, 118. It may occur in baseball and tennis players, javelin throwers and swimmers. These subjects have posterior shoulder pain that starts during the late elevating phase of overhead movement and becomes worse during the early speeding up phase119. MRI might be helpful and may reveal tears and degeneration of the posterior undersurface of the rotator cuff (supra- and infraspinatus), defects in posterior-superior labrum (SLAP IIB lesion), subcortical cysts at the humeral head, anterior capsule laxity, instability and posterior capsule enlargement119-121. It has been proposed that presence of anterior laxity of the capsule in these patients has been proposed that this is the primary problem especially in athletes who engage in overhead activities117. Anterior-superior Impingement Anterior-superior impingement is induced when the deep surface of subscapularis tendon and the common humeral insertion of the superior glenohumeral and coracohumeral ligaments (the reflective pulley) are impinged between the humeral head and the anterior-superior edge of the glenoid. Habermeyer et al122 evaluated patients with this condition during arthroscopy. They found tears at the subscapularis undersurface and “reflective pulley” lesions, defects of the long head of biceps tendon, and partial tears of the supraspinatus tendon. This condition may develop when the arm is horizontally adducted, maximally internally rotated and to a varying extent elevated anteriorly123. Anterior impingement Anterior impingement occurs in younger patients with typical subacromial impingement symptoms. Arthroscopic evaluation reveals ragged tendon fibers of the rotator cuff between the superior humeral head and the anterior superior labrum124. Entrapment of the long head of biceps tendon Patients suffering from entrapment of the long head of biceps tendon have persistent anterior shoulder pain which increases during active elevation of the arm above the head. In this condition the intraarticular portion of the long head of the biceps tendon becomes enlarged and adopts the shape of an hour glass125. When the shoulder is elevated it does not succeed to enter the biceps groove and the tendon becomes squeezed resulting in pain and limitations of motion. The place for the biceps groove is tender and at passive elevation there is a reduction of 10-20° compared to the opposite side. Arthroscopic evaluation gives the final diagnosis. The “hourglass test” is performed where the incarceration and bulging of the tendon is seen when 22 passive elevation of the arm with the elbow extended is performed. MR and CT scans are of minor diagnostic value125. Impingement not primarily related to shoulder joint anatomy Muscle weakness Because of weakness of the rotator cuff muscles tension overload may occur resulting in pathological changes in the supraspinatus tendon126. This happens when the arm is in the overhead position. These conditions are commonly seen in athletics that swim or participate in racquet or throwing sports. Manual labor that requires overhead motions such as those performed by carpenters mechanics, plumbers, and other works might also affected. Muscle fatigue, injury and degenerative changes in tendons127, 128 have been associated with proximal migration of the humeral head. Jerosch et al128 studied 8 cadavers and concluded that impingement could be caused by muscle imbalance. Consequently, these authors proposed that impingement should be treated with muscle-strengthening exercises rather than acromioplasty. Overuse of the Shoulder The overuse syndrome, which is based on repetitive overhead motions, is another reason for tendinitis, bursitis, and impingement129, 130. The overuse syndrome commonly occurs in young competitive athletes who perform forceful repetitive tasks that involve overhead motion. The most common of these activities include throwing, racquet sports, and swimming. The balance of the forces of the shoulder can be disturbed by negligible changes in the technique that an athlete uses to perform a motion, exceeding tolerance level of the soft tissue with an injury as a consequence. Inflammation and thickening of the rotator cuff tendons or the subacromial bursa may develop into subacromial impingement. The primary cause is an overuse of the shoulder and as a consequence soft-tissue inflammation. This increases the volume of the cuff tendon and bursa in the subacromial space and impingement against the coracoacromial arc129, 131, 105. A variety of diseases e.g. rheumatoid arthritis can induce inflammation with increasing volume of the rotator cuff and the bursa. Degenerative tendinopathy In a radiographic and histological study of 76 cadaveric shoulders, Ogata and Uhthoff132 showed that degenerative tendinopathy may play an important role in impingement syndrome. Those authors evaluated the degenerative changes that they came across on the undersurface of the acromion. They proposed that tendon degeneration is the primary reason for partial tears of the rotator cuff. They suggested that proximal migration of the humeral head occurs when there is a partial tear, resulting in impingement and over the time develops to a complete tear. Glenohumeral Instability It is important to exclude glenohumeral instability especially in young competitive athletes with symptoms of impingement133. This condition might be one reason why these patients do not recover after an anterior acromioplasty7, 133, 134. Glenohumeral instability might also be difficult to differentiate from other intra-articular reasons for impingement such as the anterior, anterior-superior or posterior-superior types. Rotator cuff tears The normal rotator cuff is 10-12 mm wide. Partial tears have been classified82 depending on their depth into 3 grades (less than 3 mm, 3-6 mm, more than 6 mm). Neer classified the 23 involvement of the rotator cuff into three stages: Type I - inflammation without any tear, Type II - partial tear and Type - III full thickness tear. Full thickness tear has been further divided into three degrees depending on size (less than 1 cm, 1 up to less than 3 cm, more than 3 cm). An avulsion type of partial tear of the articular surface of supraspinatus at its insertion on the greater tuberosity should be considered when young athletic patients are complaining of shoulder pain83. The supraspinatus tendon is mostly involved in developing rotator cuff tears probably because of impingement by the subacromial spurs and degenerative changes of the AC-joint. A posterior-superior internal impingement has also been described in which the posterior humeral head contacts the posterior glenoid during abduction with external rotation. The tears of the supraspinatus tendon are in the majority of cases located anteriorly and may develop further anteriorly or/and posteriorly. The rotator interval, the capsule and the coracohumeral ligament are involved and the cranial portion of subscapularis tendon may also be engaged. If the tears expand further the infraspinatus tendon may also be involved. In patients with symptoms of shoulder impingement signs of biceps tendon injury should also be looked for. There might be tears of the superior fibers of the subscapularis tendon and the anterior fibers of the supraspinatus tendon resulting in a subluxation of the biceps tendon out of the intertubercular groove. As mentioned previously, the subscapularis may also be injured when tears of the supraspinatus extend anteriorly. The subscapularis tendon may also be damaged after traumatic glenohumeral dislocation or as a consequence of subcoracoid impingement. When the coracohumeral interval is narrowed impingement of the subscapularis tendon may occur84. An avulsion of the portion of the lesser tuberosity is commonly seen because tears of the subscapularis tendon usually take places near or at its insertion. Diagnosis Symptoms Pain is the most frequent symptoms in subacromial impingement. As a consequence of pain stiffness and weakness may develop. When the pain subsides, the stiffness and weakness should to a great extent disappear. If the weakness persists other diagnoses such as cervical radiculitis or entrapment of the suprascapular nerve should be considered. If the stiffness persists, frozen shoulder, inflammatory arthritis and calcific tendinitis might be present. Pain should be analyzed with respect to localization, quality, persistence, appearance and association to activity. The subacromial impingement is characterized by increasing pain, especially when working with the arm elevated corresponding to the range of motion described in the painful arc test106. Quite often, acute traumatic disorders such as bursitis may not completely resolve and may develop into impingement lesion with a persistent procedure. Therefore, the patients quite often bring to mind a specific occasion as a cause of the symptoms. Specific shoulder test The impingement sign The impingement sign (Figure 6), as described by Neer1, is performed by standing behind the patient and passively elevating the arm in the scapular plane with one hand, while the other is stabilizing the scapula. The impingement test The impingement test as described by Neer1 can be a useful instrument in the diagnosis of impingement (Neer sign, Figure 6). After sterile injection of local anesthetics into the 24 subacromial space, the test is repeated. The injection should relive the pain in case of impingement. Painful arc test Kessel and Watson introduced the “painful arc syndrome” (Figure 7) in 1977106. This is a painful position of the shoulder joint between 60 and 120° during active abduction of the arm, which indicates a disorder of the subacromial region. It should be distinguished from increasing pain up to full abduction, which is regarded as a sign of a disorder in the acromial- clavicular joint. Hawkins sign Hawkins and Kennedy100 proposed that pain during internal rotation of the arm after passive elevation of the arm to 90° as a diagnostic test of impingement (Figure 8). Figure 6. Neer sign Figure 7. Painful arc test Figure 8. Hawkins sign The apprehension test and relocation test Young patients could have impingement caused by slight glenohumeral instability. Consequently, the apprehension test and the relocation test described by Jobe135 also should also be performed and especially when younger patients seek medical attention because of shoulder pain. The apprehension test is performed in the supine position with the involved shoulder in 90° of abduction. The arm is externally rotated beyond 90°. The test is positive when the patient is apprehensive, because the humeral head begins to dislocate anteriorly95, 135. The relocation test is then performed by directing a posterior force on the proximal aspect of the humerus, thereby relieving the sensation of apprehension95, 135. The lift-off-test Disorders of the subscapularis tendon is evaluated by the lift off test where the elbow is in 90° of flexion and the arm in maximum internal rotation behind the back. The patient is then asked to lift the arm from the back. The test is considered positive if it is not possible to lift the arm from the back136. 25 The body cross test Disorders of the acromion-clavicular joints are associated with pain by direct palpation of the joint. Internal rotation of the extended arm and adduction of the arm across the chest may also elicit pain (body cross test). However, these maneuvers may also cause impingement in the subacromial space and therefore may not be specific for the identification of disorders of the A-C joint. To more specifically identify the cause of symptoms, selective injections into the both the AC-joint and the subacromial bursa are helpful137. Posterior-superior impingement test With the patient supine, posterior pain should occur when the arm is abducted 90 -110° and then externally rotated maximally. Anterior-superior impingement test Anterior pain occurs when the arm is horizontally adducted, maximally internally rotated, and anteriorly elevated to varying extents123. Radiographic Evaluation Ordinary AP radiographs could demonstrate areas of sclerosis or spur formation on the anterior edge of the acromion with corresponding areas of subcondral cysts or sclerosis of the greater tuberosity138, 139. Other differential diagnoses such as osteoarthritis of the AC or glenohumeral joints, tendinitis calcarea and indirect signs of GH instability (Bankart lesion or Hill-Sachs lesion) can be identified. An anterior-inferior projection of the acromion tilted caudally 30° may be helpful to reveal spurs of the anterior edge of acromion140. Correspondingly an AP radiograph tilted 10° in the cephalic direction may facilitate visualization of inferiorly protruding osteophytes. An axillary radiographs may be needed to diagnose an unfused acromial epiphysis114. The AC-joint is also well visualized. Neer and Poppen38 introduced the supraspinatus outlet view. This is a lateral radiograph in the scapular plane with the X-ray beam directed 10° caudally. The slope of acromion and spurs adjacent to the AC-joint are visualized90. However, superimposition of osseous structures such as the thoracic spine, the ribs, the clavicle or the scapular body may jeopardize the interpretation of this view. Ultrasonography Ultrasonography may be useful to identify moderate or large full thickness tears141, 142. Presence of subacromial impingement during abduction or elevation of the arm can be diagnosed. One major advantage with ultrasonography is that dynamic studies are possible. Magnetic Resonance imaging (MRI) The ability to diagnose partial tears and small full thickness tears has increased the last decade with the use of MRI, although it remains difficult to differentiate these lesions from rotator cuff tendinitis143. Treatment of subacromial shoulder pain Non-Operative Treatment The majority of patients with impingement improve without any surgical treatment (success rate between 50-80%)1, 10, 21, 137. Treatment usually amounts to a restriction of certain activities, both at work and during leisure time, when necessary. Non-steroidal anti- 26 inflammatory medications, subacromial injections of steroids, and physical therapy programs are frequently used. Böhmer24 introduced a new type of physiotherapy with the purpose of enabling activity without causing pain and with the intention of finding the normal “shoulder rhythm” for the individual patient. The gravitational forces on the arm were removed by a sling fixed to the ceiling (Figure 9). Figure 9. Training with a sling according to Böhmer. The training program started with rotational motions and when the patient was pain-free the program continued with flexion-extension and lastly abduction-adduction exercises. Repetitive motions in the sling with minimum experience of pain were done for about 1 hour/day. Patients trained with the physiotherapist twice a week. Gradually load was added to strengthen the rotator cuff and the muscles that stabilize the scapula. Training continued for 3 – 6 months, gradually reducing supervision. Patients were encouraged to gradually engage in leisure-time activities, which could replace the training. Three lessons were given on the anatomy and function of the shoulder, pain management and ergonomics. According to Böhmer24, 144, this treatment should be successful. In their study only 8 of 150 cases with rotator cuff disease needed surgery. Morrison et al22 evaluated 616 patients who had isolated subacromial impingement syndrome. The patients received specific physical therapy that included isometric and isotonic muscle-strengthening exercise and non-steroidal anti-inflammatory drugs. After slightly more than 2 years 413 patients (67%) had a satisfactory result and 172 (28%) an unsatisfactory result. Patients who had a type I acromion were more likely to have satisfactory results than those who had a type II or III acromion. The duration of non-operative treatment is a clinical decision that should be based on the specific set of circumstances associated with the individual patient. However, on the basis of the findings of the majority of the authors, a minimum six months trial of nonoperative treatment seems to be reasonable145-147. Operative Treatment Anterior acromioplasty with resection of the coracoacromial ligament is an established method in treating subacromial pain operatively. Removal of the lateral portion of the acromion has, however been associated with complications and unacceptable results1, 81, 85, 89. Resection of the lateral clavicle is not regularly done as part of a subacromial decompression. 27 It is indicated only when this joint is tender or when there is osteoarthritis and osteophytes causing impingement. Anterior acromioplasty can be performed with use of either the open technique described by Neer1 and modified by Rockwood95 or the arthroscopic technique148. Open acromioplasty Open anterior acromioplasty was first described by Neer in 19721. Anterior acromiopalsty involves debridement of inflamed subacromial bursa, resection of the coracoacromial ligament and any spurs, resection of the anterior-inferior aspect of the acromion and resection of distal osteophytes or if indicated of the entire acromion-clavicular joint (Figure 10). This procedure was modified by Rockwood who introduced two-step acromioplasty. In this procedure the anterior edge of the acromion is cut and then the acromioplasty according to Neer is accomplished. It is necessary to suture the anterior part of the deltoid to the acromion to preserve deltoid function. Neer1 reported that out of 16 patients all but one with AC-joint osteoarthritis had a successful outcome after acromioplasty. Post and Cohen149 studied 72 patients with subacromial impingement treated with acromioplasty. After a mean of 23 months (range 5 – 48) 64 (89%) had postoperative relief of pain. Strength and range of motion were modestly improved. Hawkins at al150 evaluated 108 decompressions in patients with chronic impingement (none with rotator cuff tear) after a longer period of time (mean 5 years). 94 patients (87%) had satisfactory results. Unsatisfactory results were more common in patients with workers´ compensation. Later studies including 50-60 cases with about 1- 4 years follow-up have revealed about 75% excellent or satisfactory results151-153. Figure 10. Acromioplasty according to Neer with modification described by Rockwood. 28 Arthroscopic Subacromial Decompression In 1987 Ellman148 introduced arthroscopic subacromial decompression as an alternative to open acromioplasty (Figure 11). The following year, Gartsman et al154 evaluated open acromioplasty versus arthroscopic decompression in 7 cadavers each. They found no difference concerning the location and amount of bone resected and proposed that the two methods could be equally effective. Figure 11. Arthroscopic Acromioplasty according to Ellman. Esch et al155 studied 71 patients with subacromial impingement treated with arthroscopic subacromial decompression. After mean of 19 months follow-up of 19 months 60 (85%) patients were satisfied and 56 patients (77%) had an excellent or good result. 28 of the 71 patients were less than forty years old with a potential for non-diagnosed subtle instability. Paulos and Franklin156 reported on 66 patients with impingement syndrome who after six months of unsuccessful non-operative treatment had arthroscopic acromioplasty. After 32 months 57 patients (86%) were satisfied with the outcome of the procedure although 14 continued to have pain at night. Adolfsson and Lysholm157 evaluated 79 patients who were operated on using arthroscopic acromioplasty for the treatment of subacromial impingement syndrome. All of the patients had instability testing and diagnostic arthroscopy. The mean follow-up was 17 months (range 9-24 months). 53 patients (67%) had an excellent or good result. Roye et al158 reported the result of arthroscopic acromioplasty in 88 patients (90 shoulders) with stage II impingement syndrome. After a mean of 41 months (24-82) 46 patients (47 shoulders) had no tear of the rotator cuff (stage-IIa impingement according to Gartsman159), and 42 patients (43 shoulders) had a partial-thickness tear (stage-IIb impingement according to Gartsman159). The result was rated as satisfactory for 72 shoulders (80%) without any difference between stages IIa and b. Open versus Arthroscopic Acromioplasty Norlin et al160 compared arthroscopic with open decompression in 20 patients with 4-5 years duration of symptoms before the operation. The preoperative symptom was 5 years for the group that had operative arthroscopy and nearly 4 years for the group that had an open procedure. After 2 years the clinical results were comparable192. Van Holsbeeck161 compared 53 patients treated with open decompression and 53 with arthroscopic decompression according to their surgeon's preference. The preoperative 29 duration of symptoms was a mean 26 months in both groups but with a very wide range (2 months -14 years). After 2 years both groups had a high percentage of satisfactory results. Associated abnormalities, such as AC-joint osteoarthritis, adhesive capsulitis, calcific tendinitis and small tears of the rotator cuff did not influence the result. Lazarus et al162 retrospectively evaluated the result of acromioplasty in 68 patients (70) shoulders. Twenty-four shoulders were treated with open acromioplasty and 46 with arthroscopic acromioplasty. The treatment option was based on the preference of the surgeon. The mean time of non-operative treatment was 14 months for the patients who had an open procedure and 9 months for the patients who had an arthroscopic procedure. The follow-up was 12 months in both groups. The mean scores for the two groups were comparable, but there was a higher percentage of excellent results in the group of patients who had been operated on using an open procedure (54% compared with 42%) and a higher percentage of worse results in the group of patients who had been managed arthroscopically (28% compared with 17 %). Patients with Workers`Compensation tended to have inferior results. Physiotherapy versus Arthroscopic Acromioplasty In the purpose of evaluating the efficiency of arthroscopic surgery, a supervised exercise regime24, and placebo soft laser treatment in patients with rotator cuff disease (stage II) Brox144 performed a randomized clinical trial in 125 patients with at least 3 month of clinical symptoms. The follow-up took place after 6 month and no difference was found between the surgery and physiotherapy groups when evaluating them according to Neer score. 30 Aims of the Study The overall aim of the study was to evaluate the kinematics of glenohumeral joints in healthy individuals in relation to patients suffering from subacromial impingement and also determine the effect of three different treatment options (physiotherapy, open surgery and arthroscopic surgery). The specific aims were: Study I: Shoulder Kinematics in 25 Patients with Impingement and 12 Controls  study three-dimensional motion of the shoulder joint at active abduction (painful arc test) in patients with impingement syndrome stage II and controls.  to evaluate the relative contribution of the glenohumeral joint in relation to the scapular rotation, the glenohumeral joint motion and spine at maximum active abduction of the arm.  determine the repeatability of active abduction of the shoulder joint. Study II: Kinematic evaluation of Neer sign and Hawkins sign  evaluate the three-dimensional motion of the shoulder joint in patients with impingement syndrome as compare to controls when placed in the Neer and Hawkins position (passive elevation and internal rotation and passive forward flexion to 90 ° in combination with internal rotation of the arm).   evaluate the relative contribution of the glenohumeral joint in relation to the scapular rotation, the glenohumeral joint motion and the spine at maximum passive abduction of the arm.   determine the repeatability of Neer sign (passive elevation with the arm internally rotated). Study III: Shoulder rhythm in patients with impingement and controls   study the relative contribution of glenohumeral motion to the total or absolute (the scapular rotation, the glenohumeral joint motion and the spine) active and passive abduction of the humerus throughout the motion and to find out if there is any difference between patients with impingement syndromes and the control group.  study whether the speed of motion (angular velocity, velocity of proximal translation of the humeral head centre) differs between those groups. Study IV: Shoulder Kinematics evaluated before and after treatment of impingement syndrome. 19 patients randomized to open surgery, arthroscopic surgery or physiotherapy   study the three-dimensional motions of the shoulder joint in patients with impingement syndrome stage II who were treated with physiotherapy, arthroscopy and open surgery.  evaluate the clinical outcome on the basis of Constant-75 score.   determine if there is correlation between changes in Constant-75 score before and after treatment and any corresponding changes of the shoulder kinematics. 31 Patients, methods and study design Patients and controls Study I: Shoulder Kinematics in 25 Patients with Impingement and 12 Controls The kinematics of the glenohumeral joint at active abduction was studied. 25 patients, 16 men and 9 women mean aged 51 years, median aged 50, range 29–63 years with shoulder symptoms (Neer Stage 2) for more than 18 months and without rotator cuff tears or osteo- arthritis were included. All patients were examined using radiography and ultrasonography Twelve healthy volunteers (controls), 8 men and 4 women (mean and median age 32, 22–59 years) without shoulder symptoms constituted the control group (Table B, appendix) Study II: Kinematic evaluation of Neer sign and Hawkins sign The kinematics in Neer sign and Hawkins sign were evaluated in 18 patients, 11 men and 7 women (median age 49, 29-65 years) with shoulder symptoms (Neer stage 2) for more than 18 months and without rotator cuff tear or osteoarthritis were included. They were recruited from a clinical study aimed to evaluate different treatments of shoulder impingement syndrome. All patients were examined with radiography and ultrasonography. Eleven volunteers without shoulder symptoms constituted a control group, 4 men and 7 women (median age 32, 22-58 years). Only one of each patient/volunteers’ shoulders was used in the study. The gender distribution between controls and patients did not statistically differ (p=0.26, Fisher's Exact Test), but the controls was significantly younger (p=0.02, Mann- Whitney Test, Table B, appendix). Study III: Shoulder rhythm in patients with impingement and controls In the evaluation of the shoulder rhythm, relative contribution of glenohumeral motion to the total or absolute active and passive abduction (the scapular rotation, the glenohumeral joint motion and the spine) of the humerus throughout the motion were studied. Thirty patients (median age 49, 29-63 years, 20 men) respectively 21 patients (median age 50, 29-63 years, 13 men) were included. Four men and 7 women (mean age 38, median 36, 22-58 years) without shoulder symptoms constituted a control group in the active abduction group and 4 men and 5 women (mean age 35, median 30, 22-58 years) without shoulder symptoms constituted a control group in the group performing passive abduction. The gender distribution between the control group and patients in the studies of active and passive abduction did not statistically differ (p=0.09, and 0.28, Fisher's Exact Test), but the individuals in the control group were younger (p= 0.008, 0.004, Mann-Whitney Test, Table B, appendix). Study IV: Shoulder Kinematics evaluated before and after treatment of impingement syndrome. 19 patients randomized to open surgery, arthroscopic surgery or physiotherapy 52 patients primarily recruited to the subgroup evaluated with RSA, 19 (13 male, 6 female, median age 51, 37 - 63 years) participated throughout the study period. All of the patients had had symptoms for at least 18 months and were examined using radiography and ultrasonography for exclusion of rotator cuff tear, osteoarthritis of generalized joint disease such as rheumatoid arthritis. In the physiotherapy group 17 patients were originally recruited. Six did not reach the relative abduction 30°-50° at one or both of the occasions. Three patients had no follow-up, 1 had a stroke and 2 did not want to participate). One patient was excluded because of poor scatter of visualized tantalum markers (high conditions number) leaving 7 patients, 6 male and 1 female (54, 39-63 years), to be studied. 32 In the open surgery group 19 patients were at first randomized to this treatment option. Two refrained from surgery due to spontaneous improvement and one after having changed his occupation. Three patients lacked complete observations between relative abduction 30°- 50° at the follow-up evaluation, in 2 patients too few markers had been visualized, 2 turned out to have rupture of the rotator cuff, one was suffering from severe heart disease and one emigrated leaving 7 patients, 5 male, 2 female (52, 39-61 years) with complete data. In the arthroscopic surgery group only 5 patients, 2 male and 3 female (54, 49-58 years) could be included. Reasons for drop out were spontaneous recovery (4), unwillingness to participate (2), and lack of complete observations 30°-50° (1), poor marker scatter of visualized tantalum spheres (1), additional surgery because of shoulder instability (1) and detection of rotator cuff injury (1). Eleven patients participated in studies with their left shoulder and 10 patients with the right shoulder (Table B, appendix). Methods Patient history and Clinical examination (Study I-IV) Patient history All of the patients were recruited from a larger study where the aim was to evaluate the result of treating patients with impingement, stage II with 3 different treatment options: physiotherapy, open surgery and arthroscopic surgery. In this evaluation the patients had a formula including questions regarding SF 36, Simple shoulder test, UCLA score and Constant score and Constant-75 score (only outcome parameter accounted for in this Thesis). The surgeon (EH) did an initial examination first and the follow-up (after 3, 6, 12, and 24 months) was done by an independent physiotherapist. These form posed questions about possible neurologic disorders, spine disorder, chronic disorders of the joints (OsteoArthritis, Rheumatoid Arthritis), beginning of symptoms, period of symptoms, nature of pain; whether it occurs during activity, at rest, at night. In addition they were asked about pain localization, strength physical condition, instability of shoulder, side of dominant arm, occupation, psychosocial situation, frequency of sick leave and instability. Clinical examination The clinical examination took place at the first consultation done by the surgeon (EH) but the follow-up (3, 6, 12 and 24 months later) was carried out by an independent physiotherapist. The examination was focused on patients with suspected subacromial impingement and shoulder disorders that might resemble this diagnosis. The examination followed the above described formula for the 4 different score system. The investigation included the cervical spine to exclude any nerve disorders or muscle disorders, possible shoulder instability by testing apprehension, and relocation test, possible chronic joint disorders by testing adjacent joints as far as range of motion goes and also other joints of the body if there was any sign of disorders. The body cross test was used for testing the AC-joint. Inspection of the shoulder then followed with the aim of finding muscle atrophy. Palpation of the glenohumeral joint, AC-joint, and sternoclavicular joint was done and any tenderness of the joints was registered. The active and passive range of motion was documented with goniometric measurements. A specific test for diagnosing subacromial impingement was used, such as Neer sign and Hawkins sign. Also the painful arc test was performed. For testing the rotator cuff, Jobe`s sign (supraspinatus muscle) and lift of sign was used. The diagnosis was confirmed with the Neer test where 10 ml Xylocain 1% or similar anaesthetic was injected in the subacromial space resulting in substantial pain relief. 33 Radiographic examination (Study I-IV) Because of this study, one the authors of this study and one of the radiologists of the Radiographic Department of Uddevalla Hospital decided to list a series of important radiographic examinations and views for the diagnosis of impingement of the shoulder and also to find any other disorders. The painful shoulders of all of the patients were examined with anterior-posterior radiographs with the shoulder in internal and external rotation. Through this examination any disorders of the glenohumeral joint including the AC-joint could be detected. Also if there were any signs of total rotator cuff tear with a minimum distance of 6 mm between the acromion and the humeral head, this could be detected. The patients were also examined by a “supraoutlet view” for the purpose of evaluating the morphology of acromion and also to determine whether there were any spurs interfering with the supra outlet area (Neer). An axillary view was also taken for the purpose of further evaluating the AC-joint. Ultrasonographic examination (Study I-IV) Every patient who took part in the study was examined by an experienced radiologist who knows ultrasonography. 2 radiologists performed the study. The opposite shoulder was also examined to determine whether there was any pathology. The radiologists were looking for partial and total tears of the rotator cuff and also observing the way the rotator cuff reacted while the patient was moving, his or her arm. Radiostereometry (Study I-IV) In our study the kinematics of the shoulder was evaluated by use of both static technique (Hawkins sign) and by dynamic technique (Painful arc test and Neer sign). Four to 6 spherical tantalum markers (Ø = 0.8 or 1.0 mm) were inserted into the scapula (acromion) and the humeral head under local anaesthesia. Two to six weeks later the patients were coming to the laboratory for performing the RSA evaluation. A set up was used in which two radiographic- tubes were attached to the ceiling in front of the two film exchangers (Figure 12). These were placed side by side and designed for simultaneous exposure. Figure 12. A set up of 2 radiographic-tubes and 2 film-exchangers, side by side designed for simultaneous exposures. The height of the tubes and the film exchangers could be adjusted according to the height of the shoulder. Each film exchanger had a reference plate facing the screen adapted with tantalum markers in a regular manner167, 169, 172. A calibration exposure was done to determine the positions of the 2 roentgen foci and the coordinates. This exposure was used as a reference 34 examination and the information from this exposure (the coordinates of the tantalum markers) was later transferred by a computer program (UmRsa) to the subsequent exposure of the patient and the calculation of the subsequent motion of the arm was referred to the reference examination61. Hawkins sign (Figure 13) was studied statically, i.e. the shoulder joint was exposed with only one double exposure when the examiner (E.H) placed the arm and shoulder in elevation and internal rotation. Figure 13. Hawkins sign. Patient placed in front of the film-exchangers. Each subject (patients and controls) performed together with one of the authors (E.H.) several trials of passive (Neer sign, Figure 14) and active (Painful arc test, Figure 15) abduction to feel as comfortable as possible before the radiographic examination were initiated and to obtain a constant speed as possible. Figure 14. Neer sign. Patient placed in front of the film-exchangers. 35 Figure 15. Painful arc test. Patient placed in front of the film-exchangers In these trials in passive elevation (Neer sign) the examiner initially stabilized the scapula with his hand until the patients themselves could maintain the scapula at as fixed a position as possible without this intervention so that the hand of the examiner would not be included in the radiographic field of view. The patients were trained to maintain the glenohumeral joint within the limits of the aperture (film size 35 x 35 cm). A starting or reference position initiated the radiographic examination. A pair of stereoradiographs was exposed with the arm aligned to the longitudinal axis of the body and the forearm in external rotation with the palm facing forward corresponding to a well-defined anatomic position. All subsequent recordings were related to this position of the arm (Figure 12). The dynamic recordings were started using a speed of two simultaneous stereo- radiographic exposures per second during five to six seconds. Due to failure to obtain exact synchronization between the speed of the film-exchanger and the motion of the arm there was always a reduction but at least 6 or more (of 10 available) representative pairs of stereographs (films) could be included in the final analysis in each study for patients. There were at least 7 corresponding representative pairs of stereographs (films) for the controls. The radiographs were scanned at 300 dpi using a flat-bed scanner (Sharp JX 610, Osaka, Japan) and measured using dedicated software173. A fictive point corresponding to the humeral head centre was constructed to enable measurements of humeral head translations in a reproducible way. Circular templates were used to find the head centre, but only on the two images of the reference position. By using the RSA digital software the position of these centres was measured on the two images and its three dimensional coordinates were computed in the same way as for a tantalum marker, see above Thus, this plotting was done once in each shoulder. Thereafter, the position of this point was transferred to all other subsequent examinations of the same shoulder using its computed position relative to the humeral head markers. Presence of documented stable and sufficiently well-scattered tantalum marker in the humeral head is a prerequisite for these computations172. We measured rotations and translations of the humeral head using the scapula as a fixed reference segment. This was managed mathematically by using reversed rotation matrix calculations (Figure 16)163. 36 Figure 16. Schematic illustration of absolute (top) and relative glenohumeral motions. In RSA rotations are calculated in a specific order: first around the transverse, thereafter around the longitudinal and finally around the anterior-posterior axis. Since the examiner elevated the patients arm corresponding to rotations around the anterior-posterior (AP) axis, we decided to adjust the position of the cage coordinate system 90° by rotation around the longitudinal axis. This means that in this study rotations were calculated in the order abduction/adduction (AP axis), internal/external rotation (longitudinal axis) and flexion/extension (transverse axis, Figure 17). Figure 17. Order of calculation of the glenohumeral rotation used in this Thesis. 37 To estimate the contribution of scapular abduction at maximum relative abduction in the glenohumeral joint, we also recorded the absolute abduction of the humerus at maximum abduction recorded as humeral rotation around the AP axis. These data were extracted from the same recording used to obtain information about the relative motions. The absolute abduction is the rotation of the humerus around the anterior-posterior axis in relation to the cage coordinate system. It is the sum of the relative abduction in the glenohumeral joint, the scapula and the trunk. The mean values for the error of rigid body fitting (mean value for each patient segment - an indicator of marker stability) were calculated in the reference segment (scapulae) and the moving segment (humerus) in each study. The mean of values indicating marker scatter (condition numbers) were also calculated in the reference segment (scapulae) and in the moving segment (humerus) in each study. The reproducibility was tested in both active abduction (Painful arc test, 6 patients) in study I and passive abduction (Neer sign, 3 patients) study II, who repeated the active respectively passive abduction after a time interval of 15 minutes. Randomization (Study IV) The patients were allocated into three treatment options: physiotherapy, open surgery and arthroscopic surgery using closed envelopes. The randomization was based on age (<55, >55) years and gender. Treatment options (Study IV) Non-Operative Treatment The patients in study IV who were randomized to physiotherapy were following the method described by Böhmer24. The purpose of the treatment was to let the patients find their normal kinematics of the shoulder, “the shoulder rhythm” without experiencing pain. The gravitational forces on the arm were removed by suspending the arm in a sling fixed to the ceiling (Figure 9). The training program started with rotational motions of the arm. As soon as the patient could perform this movement without pain the therapy went on to include flexion/extension and lastly abduction-adduction. Repetitive motions in the sling with minimum experience of pain were done for about 60 minutes every day. Patients were trained together with the physiotherapist twice a week and performed the training program by themselves, the rest of the week. Gradually load was added to strengthen the rotator cuff and the scapular stabilizing muscles. The training continued for three to six months, with the supervision gradually being reduced. Patients were encouraged to gradually engage in leisure activities that could replace the training session depending upon what kind of physical activities they were used to. At the same time as the patients got instructions for training the physiotherapist also explained the pathogenesis for the patient. The aim being that they understand why they were undergoing a specific training program. It was our intension to make the patients more motivated to do the exercises. Open acromioplasty Before surgery the patient had had his symptoms for at least 18 months and had had conventional physiotherapy and other treatments such as anti-inflammatory drugs and steroid injections in the subacromial space. Open anterior acromioplasty was first described by Neer in 19721. Anterior acromioplasty involves debridement of inflamed subacromial bursa, resection of the coracoacromial ligament and any spurs that are present, resection of the anterior-inferior aspect of the acromion, and resection of distal osteophytes from the acromioclavicular joint or of the entire joint if there is preoperative tenderness. The procedure was modified by Rockwood and Lyons95 who introduced two-step acromioplasty. In this procedure the anterior edge of the 38 acromion cuts and then the acromioplasty according to Neer is accomplished. In this procedure it is necessary to suture the anterior part of the deltoid to the acromion to preserve the deltoid function. This was not always necessary and depended on the amount of bone that was removed at the anterior edge of the acromion (Figure 10). This procedure was done on an out-patient basis. Before leaving the hospital subjects were Informed about postoperative treatment of the wound and given a training program as well as a postoperative appointment with the surgeon (one of the authors) 3 to 6 weeks later. At the same time there was a message sent by the patient to the physiotherapist with instructions on physiotherapy according to Böhmer. The patient was given an ice-pack in a sling for reducing the pain postoperatively and this function as well as a sling the first day after the operation. The patient was ordered as soon as possible to start with physical activity as soon as possible, but within the limit of pain. Arthroscopic Subacromial Decompression Prior to the surgery the patient had has his symptoms for at least 18 months and had had conventional physiotherapy as well as other treatments such as anti-inflammatory drugs and steroid injections in the subacromial space. In 1987 Ellman described arthroscopic subacromial decompression as an alternative to open acromioplasty148. In our study the arthroscopic procedure was done with the patient in a lateral decubital position. A traction device was applied to the arm and a tension to the arm corresponding 40 Newton was applied. The shoulder was in 10° of flexion and 40° of abduction. The procedure started with marketing the bony landmarks of the shoulder, acromion, the clavicle, the AC-joint and the coracoid and the coracoacromial ligament. A portal for the arthroscopy was created at the dorsal site of the shoulder one cm medial of the most lateral point of the acromion of the shoulder and two cm distal of the lateral part of acromion (“soft spot”). The glenohumeral joint is first evaluated by the arthroscopy where one is looking for cartilage changes and rotator cuff disorder. The biceps tendon and the labrum are also investigated. The subacromial space is then visualized from the same portal and a bursectomy is performed by a shaver introduced to the subacromial space by a lateral portal. A resection of the anterior edge of the acromion of about 5-8 mm is done depending on the amount of spurs according to the preoperative radiographs and what is detected by the arthroscopic procedure. Finally a resection of about 5-8 mm of the anterior third of the under space of acromion is done (Figure 11). The same procedure postoperatively was done as described after surgery with open arthroplasty Clinical evaluation (Study IV) The clinical outcome has also been evaluated by Constant-75 Score (Constant score excluding strength test174, 175) and its sub score for pain.The Constant score was preoperatively recorded by one of the authors (EH) and then after 3, 6, 12 and 24 months by an independent physiotherapist. Statistics Study I. Data for each type of motion analyzed were interpolated linearly at 5° intervals of abduction. Statistical analyses (SPSS 13.0 for Windows, Chicago, Il, USA) were based on recordings between 25° and 55° of abduction in the glenohumeral joint using scapulae as a fixed reference segment. The selection of the 25° to 55° interval was done to maximize the number of observations and meant that 25 patients and 12 controls could be included in the statistical analysis. Repeated measure ANOVA was used. The significance level was set at p <0.05. 39 In the test of repeatability pooled (= a way to average) standard deviations are presented as a simplification to account for changes of variations during the arc of motion in each individual. Study II. Statistical analyses (SPSS 13.0 for Windows, Chicago, Il, USA) were based on recordings between 20° and 55° of passive elevation in the glenohumeral joint using scapulae as a fixed reference segment. This interval was chosen to maximize the number of available observations from each group (18 patients, 11 controls). Repeated measure ANOVA (MANOVA) was used to evaluate shoulder motions during the Neer manoeuvre. Any differences of humeral head position between patients and controls in the Hawkins position were evaluated using Mann-Whitney u-test. The significance level was set at p<0.05. In the test of repeatability pooled (= a way to average) standard deviations are presented as a simplification to account for changes of variations during the arc of motion in each individual. Study III. Statistical analyses (SPSS 13.0 for Windows, Chicago, Il, USA) were based on recordings between 20° and 55° of relative active and passive abduction in the glenohumeral joint using scapulae as a fixed reference segment. This interval was chosen to maximize the number of available observations from each group, active abduction (painful arc test, 30 patients, 11 controls) and passive abduction (Neer sign, 21 patients, 9 controls) Non-parametric tests were used in evaluation where each patient contributed with one observation. Repeated measure ANOVA (MANOVA) was used when each subject contributed with a series of dependent observations. Non-parametric correlation was used. The significance level was set at p<0.05. In the test of repeatability pooled (= a way to average) standard deviations are presented as a simplification to account for changes of variations during the arc of motion in each individual. Study IV. All statistical evaluations were based on non-parametric tests because of the limited number of observations. Changes between the preoperative and follow up examinations were evaluated using the Wilcoxon signed rank test. Comparison of radiostereometric data at active abduction was done based on data recorded at 50 degrees of active abduction using Kruskal- Wallis test. The same test was used to evaluate radiostereometric data recorded for the Hawkins test and Constant75 score and its subscore for pain. All statistical evaluations were done on the observed differences. Radiostereometric data were based on differences between the studies done before and those recorded after the treatment (physiotherapy: median 38 months 14-68, arthroscopic surgery: median 27 months 26-35, open surgery median 29 months (13–34) after the treatment. The clinical data were based on differences between the first evaluation before the initiation of the treatment and the one performed 2 years later after the treatment. Stepwise linear regression analysis was used to study whether a change of Constant-75 score was associated with changes of the humeral head position at 50° of active abduction or the test for Hawkins sign. Ethics The local ethics committee at the Medical Faculty at the University of Göteborg approved both the study of patients and controls based on two separate applications (registration numbers 475-95, 520-97). 40 Results Study I: Shoulder Kinematics in 25 Patients with Impingement and 12 Controls Painful arc test (active abduction) The rotation and translation around the 3 cardinal axes in active abduction (painful arc test) in patients with impingement stage II and controls were as follows: During the initial phase of active abduction, the humerus was slightly extended compared to its reference (starting position) of the hanging arm and the palm facing forwards. With proceeding abduction, there was an associated flexion in both groups, which reached a mean of 10° at 60º abduction, irrespective of the presence of impingement or not (Figure 18). In both groups there was also similar external rotation of the humerus of about 20º and 35º in the patients and controls, respectively (Figure 19). Figure 18-19. Flexion/Extension of humerus during active abduction, Mean, SE. The “distortion” of the mean at 0-10° and 60 - 70° of abduction is partly due to missing observations (left). Internal/External rotation during abduction. Mean, SE (right). In patients and controls, the humeral head center shifted at a mean of 1 and 2 mm medially with abduction of the shoulder joint (p=0.2, Figure 20). Small proximal displacements were observed in both groups. Patients with impingement did, however, maintain a position, which was 1 to 1.5 mm more proximal (p=0.04) during the motion (Figure 21). Patients with impingement showed almost no mean AP shift of the head centre. In controls there was a minor mean anterior displacement but without any difference compared to the patient group. (p=0.3, Figure 22). 41 Figure 20-21. Medial/Lateral translation of humeral head centre during active abduction (left). Proximal/Distal translation of humeral head centre during active abduction (right). See also legend to Figure 18-19. Figure 22. Anterior-posterior translation of humeral head centre during active abduction, Mean, SE See also legend to Figure 18-19. The glenohumeral ratio at maximum absolute and total active abduction (Painful arc test) The relative contribution of the glenohumeral joint and the scapula/trunk to the abduction of the arm was about the same in patients and controls. The abduction of the arm related to the fixed coordinate system of the cage and amounted to 139°, 101°–168° and 137°, 126°–155° in patients and controls, respectively. The corresponding mean relative maximum abduction of the humerus with the scapula fixed (in the glenohumeral joint) was also similar in the 2 groups (patients: 70°, 57°–93°, controls: 72°, 42°–83°). Thus, only about 50% of the total motion occurred in the glenohumeral joint, irrespective of the presence of impingement or not. 42 Reproducibility Repeated examinations of active abduction flexion/extension varied by 3.0° and internal/external rotation by 4.2º (2 pooled standard deviations). The corresponding values for medial/lateral, proximal/distal, and anterior/posterior translations of the humeral head centre were 0.6, 0.2 and 0.2 mm, respectively. Study II: Kinematic evaluation of Neer sign and Hawkins sign The three-dimensional motions of the shoulder joint in patients with impingement syndrome and controls were studied in tests of Neer sign (passive elevation and internal rotation) and Hawkins sign (passive forward flexion to 90° in combination with internal rotation of the arm). Neer sign Rotation and translation All the patients and the controls managed to reach at least 55° relative abduction of the glenohumeral joint (mathematically fixed scapula). Initially the humerus also moved into slight flexion. With proceeding abduction up to 55°, the flexion turned to extension in both groups and reached a mean of 28°, -36°–63° in patients and a mean of 22°, -48°–42° in controls. Figure 23-24. Flexion/extension of humerus during passive elevation, Mean, SE (left). Because of partly missing observations there is a “distortion” of the mean at 0-10 and 60-70 degrees of passive elevation. Internal/external rotation during passive elevation, Mean, SE (right). (p=0.55, Figure 23, Table C, appendix). In the early phase of the motion the humerus rotated slightly internally followed by an external rotation to about a mean of 6°,-55°–35° in patients and 4°, -30°–33° in controls at 55° of relative humeral abduction (p=0.9, Figure 24, Table C, appendix). The mean proximal translation of the humeral head centre was 1.1 mm , -7–8 at 55° in patients and -0.1 mm, -4–5 in controls (p=0.4, Figure 25 , Table C, appendix). A minimum anterior translation (about 1 mm) was observed in the two groups. At 55 ° of abduction the mean anterior translation in the patient and control groups was 0.7 mm ,-7–3 and 0.9 mm, -0,4–2 respectively (p=0.9, Figure 26 , Table C, appendix). 43 Figure 25 -26. (left) Proximal/distal translation of humeral the humeral headcentre during passive elevation (left). Anterior/posterior translation of the humeral headcentre during passive elevation (right). Mean, SE. The humeral head centre translated medially both in patients (mean 1.1 mm, -6 mm – 7 mm) and controls (2.7 mm, 0 – 5 mm); p=0.12, Figure 27, Table C, appendix). Figure 27. Medial/Lateral translation of humeral headcentre during passive translation. Hawkins sign Rotation and translation In the Hawkins position the humerus showed no mean rotation into flexion or extension in the patients (mean -1°, -71° – 46°) and a tendency to flexion in the controls (mean 14°, -71°-74° 44 ,p=0.16, Table D, appendix). Internal rotation of the humerus was observed in the two groups (patients: mean 65°, 22°-106°; controls: 57°, -40° - 107°, p=0.7, Table D, appendix). The corresponding mean abduction reached 50°, -88°–16° and 71°, -128°–2°, respectively, (P=0.07, Table D, appendix). In the patient group the humeral head centre had a more lateral position than in the control group (mean -0.1 mm, -7– 7 versus 2.1 mm, -4– 4 ; p=0.02, Table D, appendix). The humeral head centre of the patients was also positioned more superiorly than in the control group (mean 3.2 mm, -1–12 vs. -1.0 mm , -21– 6; p=0.04, Table D, appendix). The anterior/posterior translations during the test did not differ (patients: mean 2.5 mm, -2–11 versus controls: mean –1.3 mm, -35 –6; p=0.07, Table D, appendix) The glenohumeral ratio to maximum absolute or total passive abduction (Neer sign) The absolute passive abduction of the humerus at maximum movement of the arm was about the same in patients and controls (patients: 151° 120–182°, controls: 158° 146–172°). The corresponding mean relative maximum abduction of the humerus with fixed scapula and trunk was also about the same (patients:67° 55–83°, controls: 66° 34–80°). Thus, only about 40% of the total motion occurred in the glenohumeral joint, irrespectively of the presence of impingement symptoms or not. Reproducibility. Repeated elevation of the arm by the same examiner was associated with a variability corresponding to 6.4° (2 pooled standard deviations) for relative flexion/extension and 9.9° for relative internal/external rotation. The corresponding values for relative medial/lateral, superior /distal, and anterior/posterior translations of the humeral head centre were 0.4, 0.5 and 0.4 mm, respectively. Study III: Shoulder rhythm in patients with impingement and controls Rotations During active abduction the patients had a reduced mobility in the glenohumeral joint up to 40         (Table E, appendix), Figure 28. Thereafter, the relative contribution of the glenohumeral joint was about the same as in controls (all observation from 20 to 55        The pattern of mobility at absolute passive abduction was rather similar in patients and controls (Table F, Figure 29, p=0.8). Figure 28-29. Absolute and relative active abduction of the humerus in the patients and the controls (left). Mean, SE. Absolute and relative passive abduction of the humerus in patients and the controls (right, p=0.8). 45 Translations Both groups showed an increasing proximal displacement of the humeral head centre with increasing active and passive abduction of the glenohumeral joint and humerus (Tables G and H (Figures 30 and 31) without any statistically significant differences between the groups (absolute active abduction: p=0.2; absolute passive abduction: p=0.1). In the control group the mean maximum absolute proximal displacement amounted to about 20 mm and in the patient group about 30 mm. The corresponding relative displacement (with fixed scapula) constituted only 0.5 and 2 mm, respectively. Figure 30-31. Absolute and relative active proximal translation of the centre of the humerus head and relative active abduction of the GH-joint in patients and the control group (left). Mean, SE. (p=0.2). Absolute and relative passive proximal translation of the centre of the humerus (right), Mean, SE. (p=0.1) Influence of age When analyzing the influence of age, in the whole material (patients and controls, 20-55º of relative abduction) there was no correlation between any of the variables recorded to describe shoulder motion and age (r = -0.1 to 0.2, p>0.4). A separate analysis only including the control group only showed that the amount of active absolute shoulder rotation increased with decreasing age (relative abduction 40 to 55º: r=0.64–0.66, p=0.04). The other parameters studied did not show any correlation to age (r = -0.3–0.6, p>0.05). Motion velocity Active abduction was initiated with angular velocity of almost 80 degrees/s in controls and 50 degrees/s in patients. It decreased with increasing abduction in both groups down to about 20 degrees/s (controls) after 3 seconds without any difference (p=0.4, Figure 32). The examiner accelerated the passive abduction of the arm for 2 seconds up to about 40-50 degrees/second followed by a decelerating motion. As expected the speed of motion between the groups was rather similar (p=0.7, Figure 33). 46 Figure 32-33. Angular velocity during active abduction. Patients with impingement versus the control group. Mean, SE. (left, P =0.4). (Right) Angular velocity during passive abduction. Patient with impingement versus control group. Mean, SE. (right, p=0.7) The speed of proximal translation during active abduction peaked 0.5–1 s earlier than the speed of proximal translation during passive abduction and maintained a more even speed of motion for about 1s, followed by a deceleration. Patients with impingement did not deviate from the control group (p=0.4, Figure 34). The speed and pattern of translation during passive abduction was rather equal to the pattern observed during active abduction and without any difference between the control and patient groups (p=0.5, Figure 35). Figure 34-35. The translation velocity during active abduction. Patients with impingement versus controls. Mean, SE. (left, p=0.4). The translation velocity during passive abduction. Patients with impingement versus controls. Mean, SE. (right, p=0.5). Reproducibility. Repeated active abduction of the arm was associated with a variability corresponding to 4° (2 pooled SD) for absolute abduction. The corresponding value for 47 absolute active proximal translation was 3.5 mm. The equivalent values for repeated passive abduction of the arm by the same examiner were 9° and 7 mm (2 pooled SD). Study IV: Shoulder Kinematics evaluated before and after treatment of impingement syndrome. 19 patients randomized to open surgery, arthroscopic surgery or physiotherapy Painful arc test (active abduction) Comparison within the group Two years after initiation of the treatment the humeral head had not changed its position at 50° of active abduction (flexion/extension, internal/external rotation; medial/lateral, proximal/distal, anterior/posterior translation: p>0.06) in two of the groups (physiotherapy, arthroscopic surgery). In the group treated with open surgery the humerus displayed about 13° increased internal rotation (p=0.03) and the humeral head centre was displaced 2 mm more laterally at 50 °of active abduction (p=0.04). Comparison between the three groups Rotation During active abduction of the shoulder joint between 30° and 50° there mean flexion up to 15° (physiotherapy group at follow up) to a mean maximum of 7-8° of extension (open surgery group – before treatment). There was also internal or external humeral rotation with mean values ranging from about 13° of internal to about 10° external rotation. The most pronounced changes and absolute maximum values were observed in the group of patients treated with open surgery. Evaluation of the changes caused by the 3 treatments revealed no differences between the treatment groups for neither flexion/extension nor internal/external rotation (p=0.29 and 0.25, Table J, appendix) Translation Between 30°and 50° of abduction the humeral head center displaced mean 0 to 2 mm medially at both the initial and the follow up examinations. Simultaneously it displaced at a mean of 1 to 3 mm proximally and the mean displacement in the anterior/posterior direction varied between +/- 2 mm. The change of the recordings at 50° of abduction performed before and after treatment did not differ between the 3 groups (medial/lateral p=0.61, proximal/distal p=0.77, anterior/posterior p=0.57, Table K, appendix). Hawkins test Comparison within groups In neither of the groups were the humeral rotations or the translations of the humeral head centre found to be influenced by the treatment (comparison of data recorded before treatment and at the follow up occasion (p!, Table L, appendix). Comparison between groups The shoulder joint rotated both into mean flexion and extension in the Hawkins position. The mean internal rotation varied between 48°and 64° before and 46° to 56° after treatment. In 48 most patients abduction was observed with mean values between 16° and 54° before and 35 to 70° after treatment in the three groups, respectively. The before/after treatment change of the rotational position of the joint during this test did not differ between the three groups (flexion/extension: p=0.71, internal/external rotation: p=0.19, abduction/adduction: p=0.23, Table L, appendix). At 30° and 50° of abduction the humeral head center displaced by a mean of 0 to 2 mm medially at both the initial and the follow up examinations. Simultaneously it displaced by a mean of 1 to 3 mm proximally and the mean displacement in the anterior/posterior direction varied between +/- 2 mm. The change of the recordings at 50° of abduction performed before and after treatment did not differ between the 3 groups (medial/lateral: p=0.28, proximal/distal: p=0.75, anterior/posterior: p=0.76, Table M, appendix). Clinical results Comparison within groups Two years after inclusion the Constant-75 score had improved from 38 to 75 (median) in the group treated with arthroscopy (p=0.04). In the other 2 groups there was no certain change (physiotherapy group: p=0.14, open surgery group: p=0.18). At evaluation of the sub score for pain no improvement was observed in the physiotherapy (p=0.23) or in the open surgery group (p=0.06), whereas the patients in the arthroscopic surgery group improved from 2 to 15 (median) points (p=0.04, Table N, appendix). Comparison between groups Patients treated with surgery improved more than those treated with physiotherapy (p=0.025, Table 13). The group treated with arthroscopy had improved with 37 points in Constant-75 score, whereas the physiotherapy group only improved with 7 points (arthroscopy vs. physiotherapy groups=0.008, Table 13. The group treated with open surgery had improved with 22 points (open surgery vs. physiotherapy: p=0.048, arthroscopic vs. open surgery: p=0.6, Table N, appendix). Evaluation of the sub score for pain revealed that the improvement (median 5, 13 and 11 for the physiotherapy, arthroscopic surgery and open surgery groups) did not differ (p=0.43, Table N, appendix). Change of motion data and change of score Regression analysis of all cases showed a tendency to more pronounced improvement of Constant-75 score with increasing shift of the humeral head centre in the medial (p=0.009, adjusted r²=0.29, p=0.02) and posterior direction at test of Hawkins sign (p=0.02, improvement of r² to 0.52). Motion data at active abduction (50°) had no influence. 49 Discussion Repeatability Both the Neer and Hawkins signs are considered highly sensitive but have poor specificity and poor predictive value176-178. The reproducibility of shoulder motions in the test of Neer sign was fairly high in our study. In accordance with Harryman et al33, we also found a large scatter of data among patients and controls, but in the individual patient the shoulder motion could be repeated with a reasonable reproducibility. According to our observations the associated flexion/extension and internal/external rotation could in the majority of cases be reproduced with a precision of less than 10° and concerning femoral head translations an upper limit of 0.5 mm was recorded. It should, however, be noted that the deviation tended to increase with increasing abduction. It is also important to emphasize that we only studied the intra-observer error. If more examiners had been involved the scatter would probably have increased. In comparison with our studies of active abduction the repeatability (precision) of rotations was poorer. Repeated active abduction could be performed with a corresponding maximum error of about 4 degrees, whereas the maximum translation error in any of the three directions was 0.6 mm. Thus it seems that the patient more accurately can repeat a defined shoulder motion than can be done by an examiner who subjects the shoulder joint to passive motions during an impingement test. Measurement error – RSA In this study we did not evaluate the error of the RSA method itself, because an absolutely fixed position of the shoulder joint could not be obtained, which is necessary to study the actual marker configurations. Based on the documented marker stability (mean error of rigid body fitting) and marker scatter (condition number) 2 standard deviations of error (95% confidence limit) for the motion parameters studied (proximal-distal translations and ab- adduction) would most probably constitute 0.25 mm and 1 degree or less respectively. This would mean that most of the variability observed could be attributed to difficulties for the patient to consistently repeat the same shoulder motion twice despite a preceding period of training. They could, however, do so with a higher reproducibility than could the examiner, when performing the passive elevation. Thus, the variability presented is partly caused by technical and partly by biological factors. Shoulder joint anatomy and kinematics Computation and presentation of three-dimensional motions may be done in different ways. The mathematical method mainly used in RSA (Euler angles) mimics the way chosen by most orthopedic examiners to describe motions. We therefore think that this method is clinically relevant even if there are alternatives. Another limitation is that data only were collected from a single and standardized series of RSA examinations only. Dynamic recordings during performance of different activities, including tasks requiring shoulder abduction above the horizontal plane would have been more relevant, but can at present not be done using RSA, for ethical and methodological reasons. Our studies may, however, provide new and basic information about shoulder joint kinematics during a standardized motion. Generally, the rotational center of the humeral head cannot be reliably determined with active or passive motion of the intact glenohumeral joint because the motion will not necessarily be purely rotational, particularly if one assumes mismatch in curvature between the glenoid and the humeral head29, 32, 179, 180. We simplified the humeral joint area configuration to a circle. This means that the point of measurement might not always have 50 been located at the center of rotation. The measuring point was, however, equal in patients and controls to minimize the risk of any systematic differences. Such differences are more likely to occur if any of the randomly located bone markers or the center of these markers have been used. Meskers et al181 found that the average radius of the humeral head is slightly larger than that of the glenoid. He assumed that the cartilage was relatively deformable, which means that in vivo the contact between the two articular areas would become closer. He, like Soslowsky et al30 declared that at least in normal shoulders the glenohumeral joint behaves like a perfect ball and socket joint, which from a methodological point of view supports our use of the head centre to represent humeral translations. Computation of the shortest distance between the humeral head and the acromion would add still more information. Such studies may become possible provided that in addition to the RSA examinations the shoulder joint is examined with computed tomography to map out the skeletal surface and at the same time is used to determine the position of the tantalum markers. Kinematics- methodology Recordings of the absolute motions are as accurate as recordings of relative motions, but they are more difficult to interpret, because they refer to a fixed coordinate system. Another and more universal problem with kinematic recording is alignment of the coordinate system to the axis of the body or a specific joint. In our study the coordinate system was aligned to the body. The transverse axis should then ideally travel through the centers of e.g. humeral heads, the longitudinal axis is vertically aligned along the humerus when placed in an anatomic position with the arm hanging along the trunk and the sagittal axis is angulated 90° in relation to these axes. To standardize our reference position we aligned the patient as well as possible to the coordinate system and all patients positioned their palms facing forwards which means that the arm and the shoulder joint was in neutral extension/flexion and abduction/adduction whereas the rotational alignment was close to maximum external rotation. During active and passive abduction most individuals initiated their motion with internal rotation probably partly because this measure implies that the activity can be performed in a more comfortable way. Kinematic evaluation of shoulder motions could be based on measurements of movements related to an external and fixed global coordinate system or “internal” coordinate systems fixed to each bone of interest. In radiostereometry tantalum markers are used to define each bone and the alignment of one of the bones to the coordinate system is defined at an arbitrarily used reference examination. During motion all bone will move in relation to the global coordinate system. These motions are labeled “absolute.” By use of inverse matrix rotations the coordinate system of one of the bones e.g. scapula might be reoriented at each position examined to its original location. By subjecting the second bone segment (e.g. the humerus) to the same inverse rotation matrix, the true relative motion of a specific joint is computed. Detailed information on the absolute motions of the humerus might, however, be of interest because they are easier to estimate through visual inspection of the patient. One problem is their interpretation, because they are the sum not only of the humeral and scapular motions, but also include any motions of the trunk. Even if the examiner tried to train the patients and the controls to avoid any such motions, this measurement cannot be expected to have been completely successful. Thus, more or less pronounced motions between the thorax and the scapula and bending of the spine are in addition to scapular and relative humero-scapular rotations included in the parameter “absolute rotation.” We focused our analysis on the glenohumeral joint rotations and translations, because symptoms are regarded to be due to shoulder joint pathology and these motions could be exactly described. 51 Shoulder motion and age Degenerative changes of the rotator cuff are supposed to increase with age182. In our study we found no such correlation in the total material. In the control group the subjects were, however, more mobile when performing an active abduction regarding rotations around the anterior-posterior axes. This finding is difficult to interpret not least because comparison of this parameter between patients and controls revealed no difference. Concerning proximal translation during active abduction, the only parameter that differed between the groups, we found no age related influence. In this study we isolated the relative abduction in the glenohumeral joint and related this mobility to the global motion of the arm, which is closer to what the examiner actually observes. This implies a simplification because out-of-plane motions are not accounted for and the contribution of different parts of the body to the absolute motions could not be mapped out in detail. For the purpose of our study we do, however, think that analysis of shoulder abduction and proximal humeral translation is of particular interest. In study I we found that the patients with impingement have significantly increased proximal translation of the humeral head at active relative abduction of the glenohumeral joint compared to a control group75. We also observed the same phenomenon when evaluating the Hawkins sign in patients with impingement 76. Abnormal shoulder motions in these patients suggest that they also have reduced, delayed, or otherwise changed synchronization of motions in the glenohumeral joint. This hypothesis could partly be confirmed in Study IV. Patient selection Some patients randomly allocated to the three treatments in study IV were excluded because of an inability to reach 50° of relative abduction corresponding to an estimated absolute motion of 105° of abduction. This could have introduced some bias in the interpretation of data. To be more inclusive we did, however, want to ascertain that active motion within the range of abduction, which causes pain, actually had occurred and therefore required observations between 30° - 50° for the each subject to be included. Another problem was that as many as 8 patients in the surgery groups spontaneously recovered partly due to the long waiting list. Even if this could be regarded as advantageous from the patient's point of view, it also implies a selection bias, probably by an overrepresentation of patients with a more severe disease. Shoulder kinematics and impingement According to our findings, external humeral rotation is coupled to active abduction of the arm. This rotation could facilitate motion of the tuberosity under the acromion, and thereby avoid impingement of soft tissues. Our observations are not quite on line with those of MacGregor183 concerning internal humeral rotation at the end of the movement. This discrepancy could partly be because many of our patients did not reach 180° of absolute motions. Another important difference is that MacGregor studied passive motions in fresh cadavers, which may not correspond to the situation in vivo. We found no significant kinematic differences between the patients and controls during the test of Neer sign. In our study the humerus was slightly flexed and internally rotated at the beginning of the motion. Extension increased with continuing abduction, whereas the early internal rotation changed to external rotation at the later part of the motion. This external rotation might explain why Flatow et al184 did not observe any subacromial rotator contact at 52 the end of the elevation in their study. The external rotation at the end of motion may place the greater tubercle away from the undersurface of acromion and thereby facilitate the motion. Harryman et al33 used cadaver shoulders and found translational motions of the humeral head. These researchers applied pressure on the humerus and the effect of any muscle activation was not studied. Chen et al185 studied 12 male volunteers (mean age, 27 years) without any shoulder symptoms using conventional radiography. They found essentially no change in the position of the humeral head as the arm was abducted from 0° to 130° (mean change = 0.3 mm). After fatigue, the motion did, however, increase to an average of 2.5 mm, perhaps because of a more distal position before the motion was initiated. We found a slight proximal displacement of about 1 mm during the early phase of abduction probably because of effect of muscular activation. With proceeding abduction there was a slight distal displacement past 20°, resulting in a position close to the one reported by Chen et al185 at maximum abduction. In accord with Deutsch et al186 and Paletta et al55 we found that presence of cuff dysfunction implied a subtle but measurable increase of the proximal shift of the humeral head. This motion had occurred before 10° of abduction. Thereafter, the head center seemed to maintain a relatively fixed position when studied in the horizontal plane. The pathophysiologic background to the observed pattern of translation in patients with impingement is unclear. It could be caused by an unknown anatomical defect or variation or changed innervations of the muscles. Another and perhaps more probable explanation could be that these patients try to stabilize their shoulders during active abduction to reduce pain. Itoi et al187 used nine fresh frozen cadaver shoulders to simulate standard clinical tests of instability, including anterior and posterior translation and sulcus tests. In all tests the anterior and inferior humeral displacements were significantly restricted if they were done with the humerus in internal but not in neutral and external rotation. Harryman et al33 postulated that humeral rotation caused asymmetric tightening of the capsule, which can result in translation of the humeral head. If the arm is internally rotated, the posterior capsule becomes tightened and pushes the head anteriorly. They called this the capsular constraint mechanism. We observed external rotation and would, according to this theory, have recorded posterior translation of the humeral head with increasing abduction. Instead, we found a minimum anterior shift, which places this theory into question. Neer sign and Hawkins sign The humeral head centre showed a consistent pattern of translation for both patients and controls in our study. It displaced superiorly, anteriorly and medially in the glenohumeral joint. Wuelker et al47 described the translations of the glenohumeral joint in eight cadaver specimens using a dynamic shoulder model during elevation. Controlled hydrodynamic actuator forces were applied to the deltoid muscle and the rotator cuff through wire cables. Using a constant force ratio, the glenohumeral joint was elevated to 90°. Translation during elevation of the glenohumeral joint between 20 ° and 90° averaged 9.0 mm superiorly and 4.4 mm anteriorly. In our study the corresponding values were smaller in patients (1.1 mm), whereas the mean value in controls was almost zero (-0.1 mm). This discrepancy may be caused by post-mortem changes in cadavers, different techniques of recording and above all presence of muscular tension in the living subjects. We found that the patients with symptoms had a significantly more lateral position of the humeral head centre when tested with Hawkins sign. The centre was also positioned more proximally than in the subjects without symptoms. The relationship between humerus and acromion during abduction and rotation of the shoulders has recently been evaluated by open MRI56, 71. In these studies the minimum acromio-humeral distance decreased significantly from 30° of abduction to 120° of abduction. The supraspinatus tendon was affected at its most 53 sensitive part at 90° of abduction and 45° of internal rotation, but the minimum acromio- humeral distance was larger than in neutral or external rotation. In our study, the head centre in the control group did not migrate superiorly, which could be interpreted as inconsistent with the above mentioned study. It should, however be noted that we did not measure the same parameter, which could explain these seemingly divergent findings. The shoulder rhythm There has been a considerable divergence in the past concerning the relative influence of scapular rotation and glenohumeral movement during abduction of the arm188, 189,41,189 stated that during coronal abduction, scapular rotation and increasing abduction of the glenohumeral joint occur simultaneously throughout most of the movement. Between 30° and 170º, the ratio of humeral to scapular rotation was claimed to be about 2 to 1. During the first 30° of abduction, they found a great variability of the scapular motions and reported translations in both the medial and lateral direction. Freedman and Munro54 reported that for every 3° of glenohumeral movement there were 2° of scapular rotation. This ratio was constant up to the final phase of abduction, when an increase corresponding to relatively more motion in the glenohumeral joint was found. Freedman and Munro54 reported mean total scapular rotation of 65° and mean total glenohumeral movement of 103°. According to our data, the relative contribution of the glenohumeral joint to the global abduction of the humerus was smaller, even though we included any rotations of the trunk. Contrary to Freedman and Munro54, we found that the latter part of arm abduction did not occur in the glenohumeral joint but rather as isolated rotation of the scapula and/or the trunk. Further only about 40-50% of the total motion occurred in the glenohumeral joint, irrespective of the presence of impingement symptoms. The reasons for these discrepancies are unclear. Different patient selection and examination techniques could 2 reasons, but it seems probable that the methods used to measure rotation and the resolution of these methods is the most important factors. We found that patients with impingement had a different distribution between absolute active abduction of the humerus and relative abduction in the glenohumeral joint. Even if the total amount of relative abduction was similar, patients tended to reduce glenohumeral abduction in the early phase of the motion. No such difference was observed during passive abduction. The reason for this difference is unknown, but could be an effect of pain. Instead of using the glenohumeral joint, which during the early phase of motion probably is more painful, patients activate their spinal and thoracoscapular muscles to benefit from bending of the spine and thereby reach the arc of motion, which is less painful. Another and perhaps less probable reason for early reduction of active glenohumeral abduction could be early degenerative changes in the acromioclavicular joint resulting in pain and secondary changes of the pattern of scapular and glenohumeral motions. Freedman and Munroe54 and Doody et al53 analyzed abduction in the scapular plane on conventional radiographs. They computed a ratio of distribution between the glenohumeral joint and thoracoscapular joint to 3:2, whereas Poppen and Walker12 measured a ratio of 5:4 after 30 degrees of abduction. In our study the relative contribution of the glenohumeral joint to the absolute active or passive abduction was smaller. In both patients and controls it constituted only about 40% during the early phase and then gradually increased to around 50% during both passive and active motion. Thus, the glenohumeral to scapula/trunk ratio was less then 1:1 during the majority of the observations. Most of the early studies monitored scapulohumeral rhythm over 45    "   variability was observed when measurements were taken at 30   s. Inman et al41 and 54 Saha180 studied abduction in the coronal plane, whereas others12, 53, 54 studied the arm elevation in the scapular plane. In a more recent analysis of arm elevation past 30° using an electro-mechanical device with a reported accuracy of about 1-2 mm and 1° GH to ST ratios lower than 2:1 was found beyond 30°, which is more consistent with our findings190. Based on our results it does, however, seem that the scapular contribution to arm elevation is greater than previously reported. Our analysis during passive abduction showed a similar pattern of the distribution of glenohumeral and scapulo-thoracic motions in patients and controls. Grachien et al58 could not demonstrate any difference in glenoid (scapula) rotation on sequential MRI images between patients with impingement and 14 controls. They studied the shoulder at 30#$ and 120  #&  &    *?        &  concerning the passive motion, but differ concerning active abduction. This difference may be because the MRI studies were done statically, whereas we exposed the shoulder joint during motion. Grachien at al191 studied the relative glenohumeral translation during active and passive abduction with three-dimensional open MRI. Fifteen healthy subjects were studied at 5 different positions of passive abduction (30°- 150°) and at 3 different positions during active abduction of the shoulder, with and without an adducting load to the arm at 60, 90 and 120 adduction. The center of the glenoid and the midpoint of the humeral head were determined by 3D reconstruction and their relative position calculated. They found that the humeral head translated inferiorly 1-2 mm at both passive and active abduction, with slightly reduced motions during muscular activity. We found that in a control group the humeral head center displaced about 1 mm proximally during early passive and active abduction and tended to be displaced slightly distally with proceeding abduction75, 76. Even if Grachien´s observations differed from ours performed during continuous shoulder motions, the magnitude of the observed displacement were within the range of 1-2 mm in both studies We found that impingement syndrome is associated with minor changes of the shoulder kinematics. These changes include an increased proximal displacement of the humeral head centre during active abduction and the test of Hawkins sign. During the test of Hawkins sign the humeral head centre also had a more lateral position in patients with impingement compared to controls. If these changes are associated with the pathogenesis of impingement syndrome clinical improvement could be expected to be associated with their disappearance, perhaps in a more or less gradual way. Findings of an association of improved Constant-75 score and a more medial shift of the humeral head centre in the Hawkins position supports this hypothesis. Effects of treatment At two years follow-up we did not find any effect of treatment on the pattern of shoulder in 2 of the groups. In patients treated with open surgery there was a significant increase of internal humeral rotation and the centre of humeral head was displaced more laterally at 50° of relative active abduction (“painful arc test”). None of these changes were consistent with normalization, which would be implying a more distal position of the humeral head centre. The etiology of the actually observed kinematic change is somewhat unclear. They might be a consequence of open treatment and surgical exposure of the subacromial space. The significance of this finding and its possible clinical implications might be weak, not at least because the comparison of the changes caused by treatment revealed no difference between the three treatment groups. 55 The interpretation of the kinematic changes occurring in patients with impingement syndrome and after treatment is difficult. It could be that shifts of the humeral head centre in the proximal/distal direction not are firmly associated to the impingement syndrome. It could also be that due to the laxity of the shoulder joint patients might to some extent adapt to such changes. More or less visible degenerative changes inside and around the shoulder joint may also result in at least some of the changes in the pattern of motion. If so, such changes can be expected to remain and especially after conservative or less invasive surgical treatment using arthroscopy. Another problem is the comparably small sample sizes resulting in low statistical significance. We also found a comparatively high intra-individual variability and especially when recording rotations in testing of the Neer sign. This variability is probably an effect of the high degree of multiplanar mobility of the shoulder joint. During a test of passive mobility it seems to be more difficult to move the arm and simultaneously avoid changes between repeated examinations of concomitant rotations around the longitudinal and transverse axis of the arm compared to the situation at active abduction. This means that it becomes difficult to statistically verify small changes of rotations. To what extent such small changes are of clinical importance is uncertain. According to our findings of association between improvements in Constant-75 score and shift of the humeral head in the Hawkins position, it seems that this could be the case at least concerning humeral head translations. The patients in this study were recruited from a prospective randomized clinical study to evaluate any clinical differences between three treatment options. When evaluated with Constant-75, the patients treated with surgery improved significantly more than those treated with physiotherapy. This result differs from a study by Brox144 who compared arthroscopic surgery and physiotherapy. The follow-up time was shorter (6 months) and more patients were included, which might explain the different outcomes. The future The motions of the shoulder joint complex has a high degree of complexity because motions originate from four joints, the sterno-clavicular joint, the acromio-clavicular joint, the glenohumeral joint and the thoracoscapular joint. In this thesis dynamic radiostereometry was used for the first time to study glenohumeral motion. This method is well suited for this purpose because motions of individual joints can be studied separately. One important limitation with our set-up is, however, is that it only allows studies of certain specific activities limited to the flexibility of the radiographic set-up and the speed of exposure. In the future, use of fast digital detectors with exposure rates of 20-40/s and use of more flexible equipment may overcome some of these problems. One limitation with study IV was the limited sample size. This means that further studies are needed to draw relevant conclusions regarding the effects of different treatments on the kinematics of the shoulder joint. Further studies are also needed to more firmly confirm any association between the glenohumeral kinematics, symptoms of impingement and any effect on these symptoms by normalization of the pattern of motion caused by this disease. Other fields of interest for dynamic studies of shoulder joint kinematics are shoulder joint instability, degenerative rupture of the supraspinatus tendon, osteoarthritis and joint replacement. Such studies have the potential to facilitate the evaluation of the pathogenesis of some of these conditions, to delineate their influence on glenohumeral motion and also be of value to assess the effect of treatment. 56 Conclusions Study I: Painful arc test (active abduction) Shoulder Kinematics in 25 Patients with Impingement and 12 Controls  Patients with impingement syndrome showed only minor deviations of the glenohumeral kinematics compared to controls. There was an increased proximal shift of the humeral head centre, which occurred in the very early phase of activity.  About 50% of the total motion of the arm occurred in the glenohumeral joint, irrespective of the presence of impingement.  The repeatability of rotations and translations during the test varied between 3.0-4.2° and 0.2-0.6 mm (2 pooled SD). Study II: Kinematic evaluation of Neer sign and Hawkins sign  Shoulder kinematics did not differ between patients with impingement and controls during the Neer manoeuvre.  In the Hawkins position the humeral head centre had a more lateral and superior position in the patient group, but the relative humeral head rotations did not differ.  About 40% of the total motion occurred in the glenohumeral joint, irrespective of the presence of impingement symptoms.  The repeatability of rotations and translations during the test varied between 6.0-9.9° and 0.4-0.5 mm (2 pooled SD). Our findings do not support the theory that abnormal passive shoulder motions precede development of impingement syndrome. The reasons for abnormal humeral head translations during active abduction (Study I) and at Hawkins sign (Study II) are not known. It could be an effect of degenerative changes or a way for the patient to avoid or reduce pain. Study III: Shoulder rhythm patients with impingement and controls at painful arc syndrome and Neer`s sign (active respectively passive abduction)  The distribution of movement between the glenohumeral joint and the upper body, during both passive and active abduction of the scapula was less or equal to 1:1 in both patients and controls.  During active abduction the patients had a reduced mobility in the glenohumeral joint up to 40         . The pattern of mobility at passive abduction of the arm was rather similar in patients and controls.  In the control group, but not in the patient group, the amount of active absolute shoulder rotation increased with decreasing age. The other parameters studied did not show any correlation to age.  The angular velocity and the speed of proximal translation during active and passive abduction of the arm did not differ between patients and controls. Study IV: Kinematics evaluated before and after treatment of impingement syndrome. 19 patients randomized to open surgery, arthroscopic surgery or physiotherapy  Two years after initiation of the treatment the humeral head had not changed its position at 50° of active abduction in two of the treatment groups (physiotherapy, arthroscopic surgery). In the group treated with open surgery the humerus displayed increased internal rotation and the humeral head centre was displaced more laterally. 57  Comparison of the changes in shoulder joint kinematics (rotations and translations) caused by the 3 treatments revealed no differences during active or passive abduction or in the position used to test Hawkins sign.  Open or arthroscopic surgery resulted in a better outcome measured as Constant-75 score.  Regression analysis of all cases showed a tendency to more pronounced improvement of Constant-75 score with increasing shift of the humeral head centre in the medial and posterior direction in the test of Hawkins sign. Motion data at active abduction (50°) had no influence. 58 Acknowledgements Johan Kärrholm, professor at the Department of Orthopedics, Sahlgrenska University Hospital, my tutor, dear friend and colleague. Without sharing your scientific knowledge this thesis would not have been written. I also would like to thank you for your patience with me and all my other different projects and for supporting me at any time, regardless of season, the time of the week or day. Jón Karlsson, professor at the Department of Orthopaedics, Sahlgrenska University Hospital, my colleague and dear friend. Thank you for all scientific advice and support . Jüri Kartus, professor at the Department of Orthopaedics at Norra Älvsborg/Uddevalla County Hospital, my college and dear friend who helped me to start up and complete my thesis by support and encouragement. Jan Nowak, MD, PhD Shoulder Service, Department of Orthopedics, Uppsala University Hospital, my colleague and dear friend for support and help at the beginning of my research and for keeping in touch along the way. Lars Karlsson, MD, Head of the Department of Orthopaedics at Norra Älvsborg/Uddevalla hospital, my college and dear friend who provided time and financial support for this work in spite of budget cuts. Lars Körner, MD, PhD, associate professor at the Department of Orthopedics at Norra Älvsborg/Uddevalla Hospital, previously head of the Department of Orthopedics, colleague and dear friend who has supported and encouraged me throughout this period. Michael Svensson, MD, PhD at the department of Orthopedics Sahlgrenska/Mölndal, colleague and dear friend, for keeping in touch and for support and encouragement. Lars Ejerhed, MD, PhD at the Department of Orthopaedics at Norra Älvsborg/Uddevalla Hospital, colleague and dear friend for support and encouragement. Hans Lindahl, MD, PhD at the Department of Orthopaedics at Norra Älvsborg/Uddevalla Hospital, colleague and dear friend for support and encouragement Colleagues at the Department of Orthopedics at Norra Älvsborg/Uddevalla Hospital for supporting me in my research. Catherine Walker Bergström, for your excellent linguistic revision, despite the time of the day. Ramin Namitabar and Leif Nyström for all the help with the computer and the databases seven days a week. Catharina Kartus for excellent drawings and Eva Fermén and Karin Davidson, for great assistance and support concerning practical matters. Johan Hallström, my dear son, who has helped me with drawings and the layout of the thesis 59 Linda Johansson, for great assistance and support concerning practical matters and formalities. Erna Andreason for your interest in practical matter of science and for always helping me with different practical questions. Karin Davidsson and Eva Fermén for great assistance and support concerning practical matters. Birgitta Runze, Bita Shareghi and Atefeh Hariri at the Lundbergs laboratory for scanning and measurements of an infinite number of RSA radiographs. Ellinor Gustavsson, Åse Olsen, Eva Nöjdh and Eva Ekengren at the RSA laboratory. Anna Apelman, Tommy Andrén, AnnLouise Olsson, Lena Virta, Åsa Öhlund, Edgardo Varas My dear friends Ulrica and Göran Malmberg who despite my focus on research the last ten years, even during our vacations, have continued to celebrate holidays with us. Last but not least, I want to thank my dear family, Christina, my wife and our two sons Johan and Magnus and our daughter Cecilia, for support during my work on this thesis and for your endless patience. This thesis has been supported by grants from Ingabritt and Arne Lundbergs Research Foundation, The Swedish Research Council, The Research and Development Department at Norra Älvsborg/Uddevalla Hospital and the County Council of Western Sweden Västra Götalands Research Foundation 60 References 1. Neer CS, 2nd. Anterior acromioplasty for the chronic impingement syndrome in the shoulder: a preliminary report. J Bone Joint Surg Am. 1972 Jan;54(1):41-50. 2. Neer CS, 2nd. Impingement lesions. Clin Orthop. 1983 Mar(173):70-7. 3. Bergenudd H, Lindgarde F, Nilsson B, Petersson CJ. Shoulder pain in middle age. A study of prevalence and relation to occupational work load and psychosocial factors. Clin Orthop Relat Res. 1988 Jun(231):234-8. 4. van der Windt DA, Koes BW, de Jong BA, Bouter LM. Shoulder disorders in general practice: incidence, patient characteristics, and management. Ann Rheum Dis. 1995 Dec;54(12):959-64. 5. van der Windt DA, Koes BW, Boeke AJ, Deville W, De Jong BA, Bouter LM. Shoulder disorders in general practice: prognostic indicators of outcome. Br J Gen Pract. 1996 Sep;46(410):519-23. 6. Vecchio P, Kavanagh R, Hazleman BL, King RH. Shoulder pain in a community- based rheumatology clinic. Br J Rheumatol. 1995 May;34(5):440-2. 7. Fu FH, Harner CD, Klein AH. Shoulder impingement syndrome. A critical review. Clin Orthop. 1991 Aug(269):162-73. 8. Chard MD, Sattelle LM, Hazleman BL. The long-term outcome of rotator cuff tendinitis--a review study. Br J Rheumatol. 1988 Oct;27(5):385-9. 9. Zuckerman JD, Kummer FJ. The influence of the coracoacromial arch anatomy on rotatorcuff tears. J Shoulder Elbow Surg. 1992;1:4-14. 10. Kamkar A, Irrgang JJ, Whitney SL. Nonoperative management of secondary shoulder impingement syndrome. J Orthop Sports Phys Ther. 1993 May;17(5):212-24. 11. Paine RM, Voight M. The role of the scapula. J Orthop Sports Phys Ther. 1993 Jul;18(1):386-91. 12. Poppen NK, Walker PS. Normal and abnormal motion of the shoulder. J Bone Joint Surg Am. 1976 Mar;58(2):195-201. 13. Roach KE, Budimen-Mak E. Arthritis Care and Research. 1991;4:143-9. 14. Solem-Bertoft E, Thuomas KA, Westerberg CE. The influence of scapular retraction and protraction on the width of the subacromial space. An MRI study. Clin Orthop. 1993 Nov(296):99-103. 15. Tibone JE, Shaffer B. Shoulder pain: when is it impingement ? J Musculo Med. 1995;12(4):65-77. 16. Warner JJ, Micheli JL. Scapulothoracic motion in normal shoulders and shoulders with gleonhumeral instability and impingement syndrome. A study using Moire topographic analysis. Clin Orthop. 1992;285:191-9. 17. Budoff JE, Nirschl RP, Guidi EJ. Debridement of partial-thickness tears of the rotator cuff without acromioplasty. Long-term follow-up and review of the literature. J Bone Joint Surg Am. 1998 May;80(5):733-48. 18. Järvholm U, Styf J, Suurkula M, Herberts P. Intramuscular pressure and muscle blood flow in supraspinatus. Eur J Appl Physiol Occup Physiol. 1988;58(3):219-24. 19. Michener LA, McClure PW, Karduna AR. Anatomical and biomechanical mechanisms of subacromial impingement syndrome. Clin Biomech (Bristol, Avon). 2003 Jun;18(5):369-79. 20. Bigliani LU, Levine WN. Subacromial impingement syndrome. J Bone Joint Surg Am. 1997 Dec;79(12):1854-68. 21. Bartolozzi A, Andreychik D, Ahmad S. Determinants of outcome in the treatment of rotator cuff disease. Clin Orthop Relat Res. 1994 Nov(308):90-7. 61 22. Morrison DS, Frogameni AD, Woodworth P. Non-operative treatment of subacromial impingement syndrome. J Bone Joint Surg Am. 1997 May;79(5):732-7. 23. McClure PW, Bialker J, Neff N, Williams G, Karduna A. Shoulder function and 3- dimensional kinematics in people with shoulder impingement syndrome before and after a 6-week exercise program. Phys Ther. 2004 Sep;84(9):832-48. 24. Böhmer AS. Trening ved kroniske degenerative skulderlidelser. Fysioterapeuten 1984;51:192 - 7. 25. Virta L, Mortenssen M, Eriksson R, Möller M. How many patients with subacromial impingement syndrome recover with physiotherapy? A follow-up study of a supervised exercise programme. Advances in Physiotherapy. 2008:1-8. 26. Järvholm U, Palmerud G, Kadefors R, Herberts P. The effect of arm support on supraspinatus muscle load during simulated assembly work and welding. Ergonomics. 1991 Jan;34(1):57-66. 27. Wuelker N, Wirth CJ, Plitz W, Roetman B. A dynamic shoulder model: reliability testing and muscle force study. J Biomech. 1995 May;28(5):489-99. 28. Rahme H, Solem-Bertoft E, Westerberg CE, Lundberg E, Sorensen S, Hilding S. The subacromial impingement syndrome. A study of results of treatment with special emphasis on predictive factors and pain-generating mechanisms. Scand J Rehabil Med. 1998 Dec;30(4):253-62. 29. Iannotti JP, Gabriel JP, Schneck SL, Evans BG, Misra S. The normal glenohumeral relationships. An anatomical study of one hundred and forty shoulders. J Bone Joint Surg Am. 1992 Apr;74(4):491-500. 30. Soslowsky LJ, Flatow EL, Bigliani LU, Mow VC. Articular geometry of the glenohumeral joint. Clin Orthop. 1992 Dec(285):181-90. 31. Kumar VP, Balasubramaniam P. The role of atmospheric pressure in stabilising the shoulder. An experimental study. J Bone Joint Surg Br. 1985 Nov;67(5):719-21. 32. Saha AK. Dynamic stability of the glenohumeral joint. Acta Orthop Scand. 1971;44:491. 33. Harryman DT, 2nd, Sidles JA, Clark JM, McQuade KJ, Gibb TD, Matsen FA, 3rd. Translation of the humeral head on the glenoid with passive glenohumeral motion. J Bone Joint Surg Am. 1990 Oct;72(9):1334-43. 34. Howell SM, Galinat BJ, Renzi AJ, Marone PJ. Normal and abnormal mechanics of the glenohumeral joint in the horizontal plane. J Bone Joint Surg Am. 1988 Feb;70(2):227-32. 35. Howell SM, Galinat BJ. The glenoid-labral socket. A constrained articular surface. Clin Orthop Relat Res. 1989 Jun(243):122-5. 36. Saha AK. The classic. Mechanism of shoulder movements and a plea for the recognition of "zero position" of glenohumeral joint. Clin Orthop Relat Res. 1983 Mar(173):3-10. 37. O'Brien SJ, Schwartz RS, Warren RF, Torzilli PA. Capsular restraints to anterior- posterior motion of the abducted shoulder: a biomechanical study. J Shoulder Elbow Surg. 1995 Jul-Aug;4(4):298-308. 38. Neer CS, 2nd, Poppen NK. Supraspinatus outlet. Orthop Trans. 1987;11:234. 39. Glousman R, Jobe F, Tibone J, Moynes D, Antonelli D, Perry J. Dynamic electromyographic analysis of the throwing shoulder with glenohumeral instability. J Bone Joint Surg Am. 1988 Feb;70(2):220-6. 40. Howell SM, Galinat BJ. The glenoid-labral socket. A constrained articular surface. Clin Orthop. 1989 Jun(243):122-5. 41. Inman VT, Saunders Mea. Observations on the function of the shoulder joint. J Bone Joint Surg Am. 1944;16:1. 62 42. Nuber GW, Jobe FW, Perry J, Moynes DR, Antonelli D. Fine wire electromyography analysis of muscles of the shoulder during swimming. Am J Sports Med. 1986 Jan- Feb;14(1):7-11. 43. Poppen NK, Walker, P.S. Forces at the glenohumeral joint in abduction. Clin Orthop Rel Res. 1978;135:165-70. 44. Warner JJ, Micheli LJ, Arslanian LE, Kennedy J, Kennedy R. Patterns of flexibility, laxity, and strength in normal shoulders and shoulders with instability and impingement. Am J Sports Med. 1990 Jul-Aug;18(4):366-75. 45. Thompson WO, Debski RE, Boardman ND, 3rd, Taskiran E, Warner JJ, Fu FH, et al. A biomechanical analysis of rotator cuff deficiency in a cadaveric model. Am J Sports Med. 1996 May-Jun;24(3):286-92. 46. Debski RE, McMahon PJ, Thompson WO, Woo SL, Warner JJ, Fu FH. A new dynamic testing apparatus to study glenohumeral joint motion. J Biomech. 1995 Jul;28(7):869-74. 47. Wuelker N, Schmotzer H, Thren K, Korell M. Translation of the glenohumeral joint with simulated active elevation. Clin Orthop Relat Res. 1994 Dec(309):193-200. 48. Meskers CG, van der Helm FC, Rozing PM. The size of the supraspinatus outlet during elevation of the arm in the frontal and sagittal plane: a 3-D model study. Clin Biomech (Bristol, Avon). 2002 May;17(4):257-66. 49. Van der Helm FCT PG. Three-dimensional recording and description of motions of the shoulder mechanism. J Biomech Eng. 1995;117:27-40. 50. Billuart F, Devun L, Gagey O, Skalli W, Mitton D. 3D kinematics of the glenohumeral joint during abduction motion: an ex vivo study. Surg Radiol Anat. 2007 Jun;29(4):291-5. 51. Crosbie J, Kilbreath SL, Hollmann L, York S. Scapulohumeral rhythm and associated spinal motion. Clin Biomech (Bristol, Avon). 2008 Feb;23(2):184-92. 52. Codman EA. The Shoulder. Rupture of the Supraspinatus Tendon and Other Lesions in or about the Subacromial Bursa Brooklyn NY: G. Miller and company, Medical Publischer, INC; 1934. 53. Doody SG, Freedman L, Waterland JC. Shoulder movements during abduction in the scapular plane. Arch Phys Med Rehabil. 1970 Oct;51(10):595-604. 54. Freedman L, Munro RR. Abductiion of the arm in the scapular plane: scapular and glenohumeral movements. A roentgenographic study. J Bone Joint Surg Am. 1966;48(8):1503-10. 55. Paletta GA, Jr., Warner JJ, Warren RF, Deutsch A, Altchek DW. Shoulder kinematics with two-plane x-ray evaluation in patients with anterior instability or rotator cuff tearing. J Shoulder Elbow Surg. 1997 Nov-Dec;6(6):516-27. 56. Graichen H, Bonel H, Stammberger T, Englmeier KH, Reiser M, Eckstein F. Subacromial space width changes during abduction and rotation--a 3-D MR imaging study. Surg Radiol Anat. 1999;21(1):59-64. 57. Graichen H, Bonel H, Stammberger T, Haubner M, Rohrer H, Englmeier KH, et al. Three-dimensional analysis of the width of the subacromial space in healthy subjects and patients with impingement syndrome. AJR Am J Roentgenol. 1999 Apr;172(4):1081-6. 58. Graichen H, Stammberger T, Bonel H, Wiedemann E, Englmeier KH, Reiser M, et al. Three-dimensional analysis of shoulder girdle and supraspinatus motion patterns in patients with impingement syndrome. J Orthop Res. 2001 Nov;19(6):1192-8. 59. Graichen H, Hinterwimmer S, von Eisenhart-Rothe R, Vogl T, Englmeier KH, Eckstein F. Effect of abducting and adducting muscle activity on glenohumeral 63 translation, scapular kinematics and subacromial space width in vivo. J Biomech. 2005 Apr;38(4):755-60. 60. Ebaugh DD, McClure PW, Karduna AR. Three-dimensional scapulothoracic motion during active and passive arm elevation. Clin Biomech (Bristol, Avon). 2005 Aug;20(7):700-9. 61. Kärrholm J. Roentgen stereophotogrammetry. Review of orthopedic applications. Acta Orthop Scand. 1989 Aug;60(4):491-503. 62. Mollier S. Ueber die Statik und Mechanik des menslichen Schulergurtels unter normalen und pathologischen Verhaltnissen. Festschr. f.c v. Kupfer.Jena. 1899. 63. Shiino K. Schultergelenkbewegungen und Schultermuskelarbeiten. Arch Anat Physiol, Suppl Anat Abtlg. 1913:1-88. 64. Hvorslev CM. Studien ueber die Bewegungen der Schulter. ArchfPhysiol (separatabdruck). 1927. 65. Lippitt S, Matsen F. Mechanisms of glenohumeral joint stability. Clin Orthop Relat Res. 1993 Jun(291):20-8. 66. Turkel SJ, Panio MW, Marshall JL, Girgis FG. Stabilizing mechanisms preventing anterior dislocation of the glenohumeral joint. J Bone Joint Surg Am. 1981 Oct;63(8):1208-17. 67. Johnson GR, Stuart PRea. A method for the measurement of three dimensional scapular movement. Clin Biomechanic. 1993;8:269-73. 68. Kapandji IA. The shoulder. Clin Rheum Dis. 1982 Dec;8(3):595-616. 69. Bagg SD, Forrest WJ. A biomechanical analysis of scapular rotation during arm abduction in the scapular plane. Am J Phys Med Rehabil. 1988 Dec;67(6):238-45. 70. Koepfler JW. Moire topography in medicine. J Biol Photogr. 1983 Jan;51(1):3-10. 71. Graichen H, Bonel H, Stammberger T, Heuck A, Englmeier KH, Reiser M, et al. A technique for determining the spatial relationship between the rotator cuff and the subacromial space in arm abduction using MRI and 3D image processing. Magn Reson Med. 1998 Oct;40(4):640-3. 72. Graichen H, Bonel H, Stammberger T, Heuck A, Englmeier KH, Reiser M, et al. [An MR-based technique for determination of the subacromial space width in subjects with and without shoulder muscle activity]. Z Orthop Ihre Grenzgeb. 1999 Jan- Feb;137(1):2-6. 73. Graichen H, Stammberger T, Bonel H, Haubner M, Englmeier KH, Reiser M, et al. Magnetic resonance-based motion analysis of the shoulder during elevation. Clin Orthop. 2000 Jan(370):154-63. 74. Pappas GP, Blemker SS, Beaulieu CF, McAdams TR, Whalen ST, Gold GE. In vivo anatomy of the Neer and Hawkins sign positions for shoulder impingement. J Shoulder Elbow Surg. 2006 Jan-Feb;15(1):40-9. 75. Hallström E, Kärrholm J. Shoulder kinematics in 25 patients with impingement and 12 controls. Clin Orthop Relat Res. 2006 Jul;448:22-7. 76. Hallström E, Kärrholm J. Kinematic evaluation of the Hawkins and Neer sign. J Shoulder Elbow Surg. 2008 Jan-Feb;17(1 Suppl):40S-7S. 77. Badley EM, Tennant A. Changing profile of joint disorders with age: findings from a postal survey of the population of Calderdale, West Yorkshire, United Kingdom. Ann Rheum Dis. 1992 Mar;51(3):366-71. 78. Bjelle A, Hagberg M, Michaelson G. Occupational and individual factors in acute shoulder-neck disorders among industrial workers. Br J Ind Med. 1981 Nov;38(4):356-63. 79. Bjelle A. Epidemiology of shoulder problems. Baillieres Clin Rheumatol. 1989 Dec;3(3):437-51. 64 80. Nygren A, Berglund A, von Koch M. Neck-and-shoulder pain, an increasing problem. Strategies for using insurance material to follow trends. Scand J Rehabil Med Suppl. 1995;32:107-12. 81. Diamond B. The obstructing Acromion: Underlying Diseases, Clinical Development, and Surgery. Springfield, Illinois; 1964. 82. Ellman H. Diagnosis and treatment of incomplete rotator cuff tears. Clin Orthop Relat Res. 1990 May(254):64-74. 83. Tuite MJ, Turnbull JR, Orwin JF. Anterior versus posterior, and rim-rent rotator cuff tears: prevalence and MR sensitivity. Skeletal Radiol. 1998 May;27(5):237-43. 84. Gerber C, Terrier F, Zehnder R, Ganz R. The subcoracoid space. An anatomic study. Clin Orthop Relat Res. 1987 Feb(215):132-8. 85. Armstrong JR. Excision of the acromion in treatment of the supraspinatus syndrome; report of 95 excisions. J Bone Joint Surg Am. 1949 Aug;31B(3):436-42. 86. Neer CS, 2nd, Marberry TA. On the disadvantages of radical acromionectomy. J Bone Joint Surg Am. 1981 Mar;63(3):416-9. 87. Meyer AW. The minuter anatomy of attrition lesions. JBone and Joint Surg, . 1931;13:341-60. 88. Goldthwait JE. An anatomic and mechanical study of the shoulder-joint, explaining many of the cases of painful shoulder, many of the recurrent dislocations, and many of the cases of brachial neuralgias or neuritis. Am J Orthop Surg. 1909:579-606. 89. Mc LH, Asherman EG. Lesions of the musculotendinous cuff of the shoulder. IV. Some observations based upon the results of surgical repair. J Bone Joint Surg Am. 1951 Jan;33(A:1):76-86. 90. Bigliani LU. The morphology of acromion and its relationship to rotatorcuff tears. Orthop Trans. 1986;10:228. 91. Morrison DS, Bigliani LU. The clinical significance of variations in acromial morphology. OrthopTrans. 1987;11:234. 92. Toivonen DA, Tuite MJ, Orwin JF. Acromial structure and tears of the rotator cuff. J Shoulder Elbow Surg. 1995 Sep-Oct;4(5):376-83. 93. Aoki M, Ishii, et al. Surgery of the shoulder. 1990. 94. Nicholson GP, Goodman DA, Flatow EL, Bigliani LU. The acromion: morphologic condition and age-related changes. A study of 420 scapulas. J Shoulder Elbow Surg. 1996 Jan-Feb;5(1):1-11. 95. Rockwood CA, Lyons FR. Shoulder impingement syndrome: diagnosis, radiographic evaluation, and treatment with a modified Neer acromioplasty. J Bone Joint Surg Am. 1993 Mar;75(3):409-24. 96. Zuckerman JD, Kummer FJ, Cuomo F, Greller M. Interobserver reliability of acromial morphology classification: an anatomic study. J Shoulder Elbow Surg. 1997 May- Jun;6(3):286-7. 97. Chang EY, Moses DA, Babb JS, Schweitzer ME. Shoulder impingement: objective 3D shape analysis of acromial morphologic features. Radiology. 2006 May;239(2):497-505. 98. Burns TP, Turba JE. Arthroscopic treatment of shoulder impingement in athletes. Am J Sports Med. 1992 Jan-Feb;20(1):13-6. 99. Ha'eri GB, Wiley AM. Shoulder impingement syndrome. Results of operative release. Clin Orthop Relat Res. 1982 Aug(168):128-22. 100. Hawkins RJ, Kennedy JC. Impingement syndrome in athletes. Am J Sports Med. 1980 May-Jun;8(3):151-8. 101. Penny JN, Welsh RP. Shoulder impingement syndromes in athletes and their surgical management. Am J Sports Med. 1981 Jan-Feb;9(1):11-5. 65 102. Burns WC, 2nd, Whipple TL. Anatomic relationships in the shoulder impingement syndrome. Clin Orthop Relat Res. 1993 Sep(294):96-102. 103. Soslowsky LJ, An CH, Johnston SP, Carpenter JE. Geometric and mechanical properties of the coracoacromial ligament and their relationship to rotator cuff disease. Clin Orthop Relat Res. 1994 Jul(304):10-7. 104. Sarkar K, Taine W, Uhthoff HK. The ultrastructure of the coracoacromial ligament in patients with chronic impingement syndrome. Clin Orthop Relat Res. 1990 May(254):49-54. 105. Uhthoff HK, Hammond DI, Sarkar K, Hooper GJ, Papoff WJ. The role of the coracoacromial ligament in the impingement syndrome. A clinical, radiological and histological study. Int Orthop. 1988;12(2):97-104. 106. Kessel L, Watson M. The painful arc syndrome. Clinical classification as a guide to management. J Bone Joint Surg Br. 1977 May;59(2):166-72. 107. Petersson CJ. Degeneration of the acromioclavicular joint. A morphological study. Acta Orthop Scand. 1983 Jun;54(3):434-8. 108. Watson M. The refractory painful arc syndrome. J Bone Joint Surg Br. 1978 Nov;60- B(4):544-6. 109. Paulson MM, Watnik NF, Dines DM. Coracoid impingement syndrome, rotator interval reconstruction, and biceps tenodesis in the overhead athlete. Orthop Clin North Am. 2001 Jul;32(3):485-93, ix. 110. Lo IK, Parten PM, Burkhart SS. Combined subcoracoid and subacromial impingement in association with anterosuperior rotator cuff tears: An arthroscopic approach. Arthroscopy. 2003 Dec;19(10):1068-78. 111. Friedman RJ, Bonutti PM, Genez B. Cine magnetic resonance imaging of the subcoracoid region. Orthopedics. 1998 May;21(5):545-8. 112. Giaroli EL, Major NM, Lemley DE, Lee J. Coracohumeral interval imaging in subcoracoid impingement syndrome on MRI. AJR Am J Roentgenol. 2006 Jan;186(1):242-6. 113. Mudge MK, Wood VE, Frykman GK. Rotator cuff tears associated with os acromiale. J Bone Joint Surg Am. 1984 Mar;66(3):427-9. 114. Edelson JG, Zuckerman J, Hershkovitz I. Os acromiale: anatomy and surgical implications. J Bone Joint Surg Br. 1993 Jul;75(4):551-5. 115. Bigliani LU, Norris TR, Fischer J, Neer CS, 2nd. The relationship between the unfused acromial epiphysis and subacromial impingement lesions. Orthop Trans. 1983;7:138. 116. Hutchinson MR, Veenstra MA. Arthroscopic decompression of shoulder impingement secondary to Os acromiale. Arthroscopy. 1993;9(1):28-32. 117. Walch G, Boileau P, Noel E, Donell ST. Impingement of the deep surface of the supraspinatus tendon on the posterosuperior glenoid rim: an arthroscopic study. J Shoulder and Elbow Surg,. 1992;1:238.45. 118. Jobe CM. Posterior superior glenoid impingement: expanded spectrum. Arthroscopy. 1995 Oct;11(5):530-6. 119. Giaroli EL, Major NM, Higgins LD. MRI of internal impingement of the shoulder. AJR Am J Roentgenol. 2005 Oct;185(4):925-9. 120. Kaplan LD, McMahon PJ, Towers J, Irrgang JJ, Rodosky MW. Internal impingement: findings on magnetic resonance imaging and arthroscopic evaluation. Arthroscopy. 2004 Sep;20(7):701-4. 121. Halbrecht JL, Tirman P, Atkin D. Internal impingement of the shoulder: comparison of findings between the throwing and nonthrowing shoulders of college baseball players. Arthroscopy. 1999 Apr;15(3):253-8. 66 122. Habermeyer P, Magosch P, Pritsch M, Scheibel MT, Lichtenberg S. Anterosuperior impingement of the shoulder as a result of pulley lesions: a prospective arthroscopic study. J Shoulder Elbow Surg. 2004 Jan-Feb;13(1):5-12. 123. Gerber C, Sebesta A. Impingement of the deep surface of the subscapularis tendon and the reflection pulley on the anterosuperior glenoid rim: a preliminary report. J Shoulder Elbow Surg. 2000 Nov-Dec;9(6):483-90. 124. Struhl S. Anterior internal impingement: An arthroscopic observation. Arthroscopy. 2002 Jan;18(1):2-7. 125. Boileau P, Ahrens PM, Hatzidakis AM. Entrapment of the long head of the biceps tendon: the hourglass biceps--a cause of pain and locking of the shoulder. J Shoulder Elbow Surg. 2004 May-Jun;13(3):249-57. 126. Nirschl RP. Rotator cuff tendinitis: basic concepts of pathoetiology. Instr Course Lect. 1989;38:439-45. 127. Leroux JL, Codine P, Thomas E, Pocholle M, Mailhe D, Blotman F. Isokinetic evaluation of rotational strength in normal shoulders and shoulders with impingement syndrome. Clin Orthop Relat Res. 1994 Jul(304):108-15. 128. Jerosch J, Castro WH, Sons HU, Moersler M. [Etiology of sub-acromial impingement syndrome--a biomechanical study]. Beitr Orthop Traumatol. 1989 Sep;36(9):411-8. 129. Jobe FW. Impingement problems in the athlete. Instr Course Lect. 1989;38:205-9. 130. McCann PD, Bigliani LU. Shoulder pain in tennis players. Sports Med. 1994 Jan;17(1):53-64. 131. Ark JW, Flock TJ, Flatow EL, Bigliani LU. Arthroscopic treatment of calcific tendinitis of the shoulder. Arthroscopy. 1992;8(2):183-8. 132. Ogata S, Uhthoff HK. Acromial enthesopathy and rotator cuff tear. A radiologic and histologic postmortem investigation of the coracoacromial arch. Clin Orthop Relat Res. 1990 May(254):39-48. 133. Jobe FW, Kvitne RS, Giangarra CE. Shoulder pain in the overhand or throwing athlete. The relationship of anterior instability and rotator cuff impingement. Orthop Rev. 1989 Sep;18(9):963-75. 134. Glousman RE. Instability versus impingement syndrome in the throwing athlete. Orthop Clin North Am. 1993 Jan;24(1):89-99. 135. Jobe FW, Bradley JP. The diagnosis and nonoperative treatment of shoulder injuries in athletes. Clin Sports Med. 1989 Jul;8(3):419-38. 136. Gobelet C, Gerber C. [Standardized examination of the shoulder]. Rev Med Suisse Romande. 1990 Feb;110(2):149-54. 137. Butters KP, Rockwood CA, Jr. Office evaluation and management of the shoulder impingement syndrome. Orthop Clin North Am. 1988 Oct;19(4):755-65. 138. Cone RO, 3rd, Resnick D, Danzig L. Shoulder impingement syndrome: radiographic evaluation. Radiology. 1984 Jan;150(1):29-33. 139. Gold RH, Seeger LL, Yao L. Imaging shoulder impingement. Skeletal Radiol. 1993 Nov;22(8):555-61. 140. Ono K, Yamamuro T, Rockwood CA. Use of a thirty-degree caudal tilt radiographs in the shoulder impingement syndrome. J Shoulder and Elbow Surg,. 1992;1:246-52. 141. Drakeford MK, Quinn MJ, Simpson SL, Pettine KA. A comparative study of ultrasonography and arthrography in evaluation of the rotator cuff. Clin Orthop Relat Res. 1990 Apr(253):118-22. 142. Paavolainen P, Ahovuo J. Ultrasonography and arthrography in the diagnosis of tears of the rotator cuff. J Bone Joint Surg Am. 1994 Mar;76(3):335-40. 67 143. Iannotti JP, Zlatkin MB, Esterhai JL, Kressel HY, Dalinka MK, Spindler KP. Magnetic resonance imaging of the shoulder. Sensitivity, specificity, and predictive value. J Bone Joint Surg Am. 1991 Jan;73(1):17-29. 144. Brox JI, Staff PH, Ljunggren AE, Brevik JI. Arthroscopic surgery compared with supervised exercises in patients with rotator cuff disease (stage II impingement syndrome). BMJ. 1993 Oct 9;307(6909):899-903. 145. Altchek DW, Warren RF, Wickiewicz TL, Skyhar MJ, Ortiz G, Schwartz E. Arthroscopic acromioplasty. Technique and results. J Bone Joint Surg Am. 1990 Sep;72(8):1198-207. 146. Bigliani LU, D'Alessandro DF, Duralde XA, McIlveen SJ. Anterior acromioplasty for subacromial impingement in patients younger than 40 years of age. Clin Orthop. 1989 Sep(246):111-6. 147. Ellman H, Kay SP. Arthroscopic subacromial decompression for chronic impingement. Two- to five-year results. J Bone Joint Surg Br. 1991 May;73(3):395-8. 148. Ellman H. Arthroscopic subacromial decompression: analysis of one- to three-year results. Arthroscopy. 1987;3(3):173-81. 149. Post M, Cohen J. Impingement syndrome. A review of late stage II and early stage III lesions. Clin Orthop Relat Res. 1986 Jun(207):126-32. 150. Hawkins RJ, Brock RM, Abrams JS, Hobeika P. Acromioplasty for impingement with an intact rotator cuff. J Bone Joint Surg Br. 1988 Nov;70(5):795-7. 151. Thorling J, Bjerneld H, Hallin G, Hovelius L, Hagg O. Acromioplasty for impingement syndrome. Acta Orthop Scand. 1985 Apr;56(2):147-8. 152. Sahlstrand T. Operations for impingement of the shoulder. Early results in 52 patients. Acta Orthop Scand. 1989 Feb;60(1):45-8. 153. Bjorkenheim JM, Paavolainen P, Ahovuo J, Slatis P. Subacromial impingement decompressed with anterior acromioplasty. Clin Orthop Relat Res. 1990 Mar(252):150-5. 154. Gartsman GM, Blair ME, Jr., Noble PC, Bennett JB, Tullos HS. Arthroscopic subacromial decompression. An anatomical study. Am J Sports Med. 1988 Jan- Feb;16(1):48-50. 155. Esch JC, Ozerkis LR, Helgager JA, Kane N, Lilliott N. Arthroscopic subacromial decompression: results according to the degree of rotator cuff tear. Arthroscopy. 1988;4(4):241-9. 156. Paulos LE, Franklin JL. Arthroscopic shoulder decompression development and application. A five year experience. Am J Sports Med. 1990 May-Jun;18(3):235-44. 157. Adolfsson L, Lysholm J. Results of arthroscopic acromioplasty related to rotator cuff lesions. Int Orthop. 1993;17(4):228-31. 158. Roye RP, Grana WA, Yates CK. Arthroscopic subacromial decompression: two- to seven-year follow-up. Arthroscopy. 1995 Jun;11(3):301-6. 159. Gartsman GM. Arthroscopic acromioplasty for lesions of the rotator cuff. J Bone Joint Surg Am. 1990 Feb;72(2):169-80. 160. Norlin R. Arthroscopic subacromial decompression versus open acromioplasty. Arthroscopy. 1989;5(4):321-3. 161. Van Holsbeeck E, DeRycke J, Declercq G, Martens M, Verstreken J, Fabry G. Subacromial impingement: open versus arthroscopic decompression. Arthroscopy. 1992;8(2):173-8. 162. Lazarus MD, Cuckler JM, Schumacher HR, Jr., Ducheyne P, Baker DG. Comparison of the inflammatory response to particulate polymethylmethacrylate debris with and without barium sulfate. J Orthop Res. 1994 Jul;12(4):532-41. 68 163. Selvik GA. A roentgen stereophotogrammetric method for the study of kinematics of the skeletal system. Thesis, University of Lund, Lund, Sweden. 1974. 164. Kärrholm J. Ankelfractures in children. Lund: University of Lund; 1982. 165. Thanner J, Freij-Larsson C, Kärrholm J, Malchau H, Wesslen B. Evaluation of Boneloc. Chemical and mechanical properties, and a randomized clinical study of 30 total hip arthroplasties. Acta Orthop Scand. 1995 Jun;66(3):207-14. 166. Kärrholm J, Herberts P, Hultmark P, Malchau H, Nivbrant B, Thanner J. Radiostereometry of hip prostheses. Review of methodology and clinical results. Clin Orthop. 1997 Nov(344):94-110. 167. Kärrholm J, Selvik G, Elmqvist LG, Hansson LI. Active knee motion after cruciate ligament rupture. Stereoradiography. Acta Orthop Scand. 1988 Apr;59(2):158-64. 168. Kärrholm J, Selvik G, Elmqvist LG, Hansson LI, Jonsson H. Three-dimensional instability of the anterior cruciate deficient knee. J Bone Joint Surg Br. 1988 Nov;70(5):777-83. 169. Uvehammer J, Kärrholm J, Brandsson S. In vivo kinematics of total knee arthroplasty. Concave versus posterior-stabilised tibial joint surface. J Bone Joint Surg Br. 2000 May;82(4):499-505. 170. Brandtsson S. Anterior Cruciate ligament injury. Gothenburg: University of Gothenburg; 2000. 171. Isberg J. Kinematics and laxity in the knee before and after Anterior Cruciate ligamnet reconstruction. University of Gothenburg: Gothenburg; 2008. 172. Nilsson KG, Kärrholm J, Ekelund L. Knee motion in total knee arthroplasty. A roentgen stereophotogrammetric analysis of the kinematics of the Tricon-M knee prosthesis. Clin Orthop. 1990 Jul(256):147-61. 173. Börlin N. The precision of radiostereometric measurements. Manual vs. digital measurements. J Biomech. 2002;35(1):69-79. 174. Constant CR, Murley AH. A clinical method of functional assessment of the shoulder. Clin Orthop Relat Res. 1987 Jan(214):160-4. 175. Constant CR. [Assessment of shoulder function]. Orthopade. 1991 Oct;20(5):289-94. 176. Calis M, Akgun K, Birtane M, Karacan I, Calis H, Tuzun F. Diagnostic values of clinical diagnostic tests in subacromial impingement syndrome. Ann Rheum Dis. 2000 Jan;59(1):44-7. 177. Leroux JL, Thomas E, Bonnel F, Blotman F. Diagnostic value of clinical tests for shoulder impingement syndrome. Rev Rhum Engl Ed. 1995 Jun;62(6):423-8. 178. MacDonald PB, Clark P, Sutherland K. An analysis of the diagnostic accuracy of the Hawkins and Neer subacromial impingement signs. J Shoulder Elbow Surg. 2000 Jul- Aug;9(4):299-301. 179. Maki S, Gruen T. Anthropometric studies of the glenohumeral joint. Trans Orthop Res Soc. 1976;1:173. 180. Saha AK. Theory of the shoulder mechanism: Descriptive and applied. Springfield, I.L.; 1961. 181. Meskers CGM, Vermeulen HM, al e. 3D Shoulder position measurement using a six degree of freedom electromagnetic tracking device. Clin Biomechanic. 1998;13:380- 92. 182. Rathbun JB, Macnab I. The microvascular pattern of the rotator cuff. J Bone Joint Surg Br. 1970 Aug;52(3):540-53. 183. Macgregor L. Rotation of the shoulder; a critical enquiry. Br J Surg. 1936;24:425. 184. Flatow EL, Soslowsky LJ, Ticker JB, Pawluk RJ, Hepler M, Ark J, et al. Excursion of the rotator cuff under the acromion. Patterns of subacromial contact. Am J Sports Med. 1994 Nov-Dec;22(6):779-88. 69 185. Chen SK, Simonian PT, Wickiewicz TL, Otis JC, Warren RF. Radiographic evaluation of glenohumeral kinematics: a muscle fatigue model. J Shoulder Elbow Surg. 1999 Jan-Feb;8(1):49-52. 186. Deutsch A, Altchek DW, Schwartz E, Otis JC, Warren RF. Radiologic measurement of superior displacement of the humeral head in the impingement syndrome. J Shoulder Elbow Surg. 1996 May-Jun;5(3):186-93. 187. Itoi E, Motzkin NE, Morrey BF, An KN. Contribution of axial arm rotation to humeral head translation. Am J Sports Med. 1994 Jul-Aug;22(4):499-503. 188. Duvall EN. Critical analysis of divergent views of movement at the shoulder-joint. Arch Phys Med Rehabil. 1955 Mar;36(3):149-54. 189. Hollinshead WH. Anatomy for surgeons. New York; 1958. 190. McQuade KJ, Smidt GL. Dynamic scapulohumeral rhythm: the effects of external resistance during elevation of the arm in the scapular plane. J Orthop Sports Phys Ther. 1998 Feb;27(2):125-33. 191. Graichen H, Stammberger T, Bonel H, Karl-Hans E, Reiser M, Eckstein F. Glenohumeral translation during active and passive elevation of the shoulder - a 3D open-MRI study. J Biomech. 2000 May;33(5):609-13. 192. Lindh M, Norlin R. Arthroscopic subacromial decompression versus open acromioplasty. A two-year follow-up study. Clin Orthop Relat Res. 1993 May;(290):174-6. 193. LaPrade RF, Wickum DJ, Griffith CJ, Ludewig PM. Kinematic evaluation of the modified Weaver-Dunn acromioclavicular joint reconstruction. Am J Sports Med. 2008 Nov;36(11):2216-21. 194. Richards DP, Burkhart SS, Campbell SE. Relation between narrowed coracohumeral distance and subscapularis tears. Arthroscopy. 2005 Oct;21(10):1223-8. 195. Kennedy J, Cameron A. Complete dislocation of the acromio-clavicular joint. J. Bone Surg Br. 1954, 36-B;203-8 70 Appendix Table B Randomization to A=Arthroscopic Subacromial Decompression, O=Open Acromioplastic, P=Physiotherapy (I ) Shoulder Kinematics in 25 Patients with Impingement and 12 Controls, (II) Kinematic evaluation of the Hawkins and Neer sign, (III) Shoulder Rhythm in Patients and Controls, Dynamic RSA during active(III(a)) and passive(III(p)) abduction. (IV) Shoulder Kinematics in 19 patients with impingement after arthroscopic surgery(as), open surgery(os) and physiotherapy(pt). Gender age sida rand I II III(a) III(p) age IV(pt) IV(os) IV(as) female 58 left A X X X X female 56 left O X X X X male 40 right P X X X X male 49 right A X X female 63 right P male 37 left P X X male 57 left P 60 X male 50 left A X X X X 52 X male 49 right O X X X 52 X male 34 right O X X male 68 right O female 62 left P male 76 left O male 49 left A X 79 X female 32 left P female 48 right A 52 X female right S female 53 left O X X female 54 right A X X female 57 right O 60 X male 53 left O female 28 left A X X X X female 54 left P X X X female 55 right A X X 58 X female 58 right O male 37 left P X X X X 41 X female 45 right P X X X X male 54 left O X X 56 X female 50 left A male 40 left A X X X female 31 left P male 63 left P X X X X 66 X female 39 right Ö X X X male 56 male 69 right A male 60 right P X X X X 62 X female 39 left P X X X X 42 X female 45 left O X X X 48 X male 51 left O female 53 right O male 56 left P X X X X male 62 right P X X X 66 X female 45 right A X X X male 60 right A male 49 left P X X X X 52 X female 45 left O male 49 right A male 51 left O X X 54 X female 58 left O male 52 right O X X 55 X female 42 right A male 48 left P X X male 61 right O X X X X 64 X subjects female 29 left X X X X female 52 left X X X male 40 left X X X X male 22 left X X X X male 32 left X X X X female 25 left X X X X female 51 right X X X X female 58 left X X X X male 30 left X X X X female 27 left X X X X female 49 right X X X 375 left Table C. Kinetics of Neer sign at 55 degrees of relative abduction 71 Type of motion Patients Controls Mean SEM CI-95% Median Mean SEM CI-95% Median p-value Rotations degrees Extension/flexion(-) 28 5 18 - 38 26 22 7.5 7 - 37 30 0.55 Internal/external(-) -6 5 4 - 16 -5 -4 6 -16 - 8 -4 0.9 Translations mm Anterior/posterior(-) 0.7 0.7 -0.7 - 2.1 1.3 0.9 0.25 0.4 - 1.4 0.9 0.9 Proximal/distal(-) 1.1 1.0 -0.9 - 3 0.9 -0.1 0.7 -1.5 - 1.3 -0.7 0.4 Medial/lateral(-) 1.1 0.8 -0.5 - 2.7 1.0 2.7 0.5 1.7 – 3.7 2.5 0.12 Table D. Kinetics of Hawkins sign Type of motion Patients Controls Mean SEM CI-95% Median Mean SEM CI-95% Median p-value Rotations degrees Extension/flexion(-) -1 7 -15 - 13 3.5 14 12 -9.5-37.5 23 0.16 Internal/external(-) 65 5 55 - 75 66 57 11 35 - 79 65 0.7 Adduction/abduction(-) -50 7 -36 - 64 -51 -71 10 -91-(-51) -71 0.07 Translations mm Anterior/posterior(-) 2.5 0.65 1.2 - 3.8 2.3 -1.3 3.4 -8 - 5 1,0 0.07 Proximal/distal(-) 3.2 0.7 1.8 - 4.6 2.6 -1.0 2.2 -5.3 - 3.3 0.8 0.04 Medial/lateral(-) -0.1 0.9 -1.9 - 1.7 0.04 2.1 0.4 1.3 - 2.9 2.5 0.02 Table E. Combined spine, trunk, scapular and humeral motions (absolute motions) at increasing degrees of motions inside the glenohumeral joint (relative motion). Recorded values and distribution between the absolute and relative glenohumeral motions in percent are presented at active abduction. Patients Controls Active abduction Active abduction Gleno- Spine, trunk, scapula Gleno- Spine, trunk, scapula Gleno- Spine, trunk, scapula Gleno- Spine, trunk, scapula humeral and glenohumeral humeral and glenohumeral humeral and glenohumeral humeral and glenohumeral mean mean CI-95% mean mean CI-95% mean mean CI-95% mean mean CI-95% degrees percent degrees percent 20 82 70 - 94 28 72 68 - 76 20 51 36 - 66 47 53 40 - 66 25 89 78 - 100 31 69 66 - 72 25 63 48 - 77 46 54 43 - 65 30 94 85 - 103 34 66 63 - 69 30 73 58 - 88 45 55 45 - 65 35 100 91 - 109 38 62 49 - 66 35 82 67 - 97 46 54 45 -63 40 103 95 - 111 38 62 59 - 66 40 90 77 - 103 48 52 44 - 60 45 107 100 - 114 44 56 53 - 58 45 94 82 - 106 50 50 43 - 57 50 110 103 - 117 48 52 49 - 55 50 99 88 - 110 52 48 42 - 54 55 114 107 - 121 50 50 47 - 53 55 104 94 - 114 54 46 40 - 52 65 125 120 - 131 53 47 43 - 51 73 125 118 - 132 58 42 37 - 47 70 144 138 -150 51 49 48 - 54 77 146 139 - 153 53 47 43 - 51 72 Table F. Combined spine, trunk, scapular and humeral motions (absolute motions) at increasing degrees of motions inside the glenohumeral joint (relative motions) Recorded values and distribution between the absolute and relative glenohumeral motions in percent are presented at passive abduction. Patients Controls Passive abduction Passive abduction Gleno- Spine, trunk, scapula Gleno- Spine, trunk, scapula Gleno- Spine, trunk, scapula Gleno- Spine, trunk, scapula humeral and glenohumeral humeral and glenohumeral humeral and glenohumeral humeral and glenohumeral mean mean CI-95% mean mean CI-95% mean mean CI-95% mean mean CI-95% degrees percent degrees percent 20 53 43 - 63 43 57 51 - 63 20 47 32 - 62 52 48 32 - 63 25 62 51 - 73 45 55 49 - 61 25 57 43 - 71 50 50 38 - 62 30 71 62 - 81 46 54 48 - 60 30 67 54 - 80 49 51 41 - 62 35 78 67 - 89 49 51 44 - 58 35 77 64 - 90 48 52 44 - 60 40 88 78 - 98 49 51 46 - 50 40 87 75 - 101 49 51 43 - 59 45 97 87 - 107 50 50 45 - 55 45 97 83 - 111 48 52 45 - 59 50 106 88 - 106 49 51 47 - 55 50 105 93 - 117 49 51 45 - 57 55 115 105 - 125 50 50 45 - 55 55 112 101 - 129 51 49 41 - 57 62 129 122 - 136 48 52 48 - 56 63 129 115 - 143 49 51 46 - 56 67 153 150 - 159 44 56 53 - 59 69 159 154 - 164 44 56 52 - 62 Table G . Proximal translation of the humeral head in combination with active spine, trunk, scapular and humeral motions (absolute motion) at increasing degrees of motion insida the glenohumeral joint (relative motion). Recorded values in mm proximal translations of the humeral head are presented. Patients Controls Active abduction Active abduction Relativ Absolute Relative Absolute Relative Absolut Relative Absolut GH motion prox prox GH motion Prox prox motion transl transl motion Transl transl mean mean mean CI-95% mean CI-95% mean mean mean CI-95% mean CI-95% degrees mm mm degrees mm mm 20 82 2.1 (1.0 - 3.2) 11.1 (3,4 - 18.7) 20 51 1.3 (0.6 - 2.0) 2,7 (-3,2 - 8,6) 25 89 2.3 (1.0 - 3.6) 14.0 (6.1 - 22.0) 25 63 1.3 (0.2 - 1.7) 4,8 (-1,5 - 11,1) 30 94 2.4 (1.7 - 3.0) 17.0 (8.7 - 25.6) 30 73 1.2 (0.4 - 1.5) 7,7 (-0,9 - 14,5) 35 100 2.4 (1.6- 3.2) 20.6 (12.3 - 28.8) 35 82 1.0 (0.2 - 1.9) 10,6 (2,4 - 17,6) 40 103 2.4 (1.6- 3.2) 24.0 (15.7 - 32.1) 40 90 1.0 (-0.07 - 1.9) 13,3 (5,2 - 21,4) 45 107 2.4 ( 1.5 - 3.3) 27.7 (18.8 - 36.5) 45 94 0.8 (-0.3 - 1.8) 16,3 (7,6 - 24,4) 50 110 2.2 (1.3 - 3.3) 31.2 (22.1 - 40.2) 50 99 0.6 (-0.5- 1.8) 19,5 (10,9 - 28,1) 55 109 2.1 (1.2- 3.1) 34.5 (25.8 - 43.3) 55 104 0.5 (-0.8 - 1.7) 23 (14,1 - 31,9) Table H. Proximal translation of the humeral head in combination with passive spine, trunk, scapular and humeral motions (absolute motion) at increasing degrees of motion insida the glenohumeral joint (relative motion). Recorded values in mm proximal translations of the humeral head are presented. At 55 degrees of relative motion there are missing observation at absolut proximal translations in patients and the control group. Patients Controls Passive abduction Passive abduction Relativ Absolute Relativ Absolut Relative Absolute Relativ Absolut GH rotation prox prox GH rotation Prox prox rotation transl translation rotation transl transl mean mean mean mean mean mean mean mean degrees mm CI-95% mm CI-95% degrees mm CI-95% mm CI-95% 20 53 1.5 (0.9 - 2.1) 6.4 (-0.3 - 13.0) 20 47 1.1 (0.7 - 1.9) 1.3 (-18.1- 11.4) 25 62 1.7 (1.1 - 2.3) 8.7 (2.0 - 15.4) 25 57 1.0 (0.3- 1.6) 0 (-13,4 - 16.1) 30 71 1.8 (0.6 - 3.1) 12 (5.0 - 18.9) 30 67 1.0 (0.1 - 1.8) 1.3 (13.3 - 15.9 ) 35 78 1.6 (0.3 - 3.0) 15.9 (8.1 - 23.6) 35 77 0.9 (-0.1 - 1.9) 4.7 (10.0 - 19.4) 40 88 1.7 (0.4 - 3.1) 20.2 (12.8 - 27.6) 40 87 0.7 (-0.5 - 1.8) 8.9 (-5.2 - 23.1) 45 100 1.4 (-0.1 - 2.8) 24.7 (16.9 - 32.4) 45 97 0.6 (-0.6 - 1.7) 14.2 (1.1 - 27.3) 50 106 1.3 (-0.3 - 3.0) 30.6 (23.4- 37.9) 50 105 0.2 (-1.3 - 1.4) 18.4 (3.8 - 32.9) 55 115 1.0 (-1 - 3.1) 36.1 (28.7-43.5) 55 111 0.0 (-1.3 - 1.4) 22.6 (7.1 - 38.2) 73 Table J. Flexion/extension, internal/external rotation (fixed scapulae) of patients with impingement syndrom randomised to arthroscopic surgery, opens surgery and physiotherapy after inclusion(preoperatively) and 38 months(follow up) after inclusion. Recorded values (mean and 95% CI) in degrees. P-value of the difference at 50° within and between the groups Flexion (-) Extension (+) Internal rotation (-) External rotation (+) relative physeotherapi artro - surgery open - surgery physeotherapi artro - surgery open - surgery active preoperatively Preoperatively abduction mean CI-95% mean CI-95% mean CI-95% mean CI-95% mean CI-95% mean CI-95% 30 -5 (-)11 - 1 0 (-)8 - 8 0 (-)7 - 7 12 8 - 17,0 5 (-)12 - 21 1 (-)12 - 13 35 -4 (-)10 - 2 2 (-)6- 10 1 (-)6 - 8 11 7 -16,0 3 (-)13- 20 -2 (-)15 - 11 40 -3 (-)9 - 4 4 (-)4- 12 3 (-)5- 11 10 5 - 15,0 2 (-)15 - 18 -4 (-)17 - 9 45 -1 (-)7 -5 7 (-)5 - 19 5 (-)3 - 13 9 4 - 14,0 0 (-)16 - 17 -6 (-)19 - 7 50 1 (-)8 -6 9 (-)4 - 21 8 (-)1 - 16 8 2 - 13,0 4 (-)13 - 21 -9 (-)22 - 5 Follow up 38 months Follow up 38 months 30 -17 (-)31 - (-)2 2,9 (-)25 - 31 -4 (-)6- 2 3 (-)7- 14 5 (-)12 - 22 11 3-19,0 35 -17 (-)32 - (-)3 2,6 (-)26- 31 -2 (-)8 - 3 3 (-)8 - 14 5 (-)12 - 22 10 1-18,0 40 -17 (-)32- (-)3 2,4 (-)27- 32 -1 (-)6 - 5 4 (-)8 - 15 5 (-)13 - 23 8 (-)1 - 17 45 -17 (-)32 - (-)3 2,4 (-)28 - 33 2 (-)4 - 7 4 (-)9- 16 5 (-)14 - 24 6 (-)0 - 10 50 -17 (-)31 - (-)3 2,4 (-)30 - 34 4 (-)2 - 9,0 3 (-)9 - 16 4 (-)16 - 24 4 (-)5 - 13 p value (within the 0,139 0,893 0,31 0,595 0,5 0,028 group) p value (between 0,29 0,25 the groups) Table K. Proximal/distal, medial/lateral, anterior/posterior translation (fixed scapulae) of the centre of the humeral head in patients with impingement syndrom randomised to arthroscopic surgery, opens surgery and physiotherapy after inclusion(preoperatively) and 38 months(follow up) after inclusion. Recorded values (mean and 95% CI) in mm. P-values at 50° abduction within and between the groups. Proximal(+) Distal (+) Medial(+) Lateral (-) Anterior(+) Posterior (-) relative physiotherapy artro - surgery open- surgery physiotherapy arto- surgery open - surgery physiotherapy artro - surgery open - surgery active Preoperatively Preoperatively Preoperatively abduct. mean CI-95% mean CI-95% mean CI-95% mean CI-95% mean CI-95% mean CI-95% mean CI-95% mean CI-95% mean CI-95% 30 2,2 0,8-3,6 1,6 0,3-3,1 2,4 0,8-4 1,6 0,6-2,6 1,5 (-)0,9-3,9 1,3 (-)0,1-2,7 1,7 0,7-2,7 1,1 (-)0,3-2,5 0,8 (-)0,8-2,4 35 2,1 0,7-3,5 1,6 0,3-3,2 2,4 0,8-4 1,9 0,9-2,9 1,6 (-)0,8-4 1,5 0,1-2,9 1,6 0,6-2,6 1,2 (-)0,2-2,6 0,8 (-)0,8-2,4 40 1,9 0,5-3,3 1,7 (-)0,1-3,5 2,4 0,6-4,2 1,8 0,6-3 1,6 (-)0,9-4,1 1,6 0-3,2 1,5 0,5-2,5 1,3 (-)0,3-2,9 0,8 (-)0,8-2,4 45 1,9 0,3-3,5 1,8 (-)0,2-3,8 2,4 0,4-4,4 2,1 0,7-3,5 1,8 (-)0,7-4,3 1,8 (-)0,7-4,3 1,4 0,4-2,4 1,7 (-)0,5-3,9 0,8 (-)0,6-2,2 50 1,8 0,2-3,4 1,9 (-)0,3-4,1 2,3 0,3-4,3 2,2 0,8-3,6 2 (-)0,5-4,5 2,1 0,3-3,9 1,2 0-2,4 1,8 (-)0,6-4,2 0,7 (-)0,5-1,9 follow up follow up follow up 30 1,7 0,5-2,9 2,1 0,7-3,5 2,5 0,5-4,5 0,7 (-)0,5-1,9 1,3 (-)1,1-3,7 (-)0,1 (-)1,1-0,9 0,2 (-)0,8-1,2 2,1 0,5-3,7 0,7 (-)0,3-1,7 35 1,5 0,3-2,7 1,9 0,3-3,5 2,5 0,5-4,5 0,8 (-)0,6-2,2 1,3 (-)1,1-3,7 (-)0,1 (-)1,3-1,1 0,2 (-)0,8-1,2 2 0,2-3,8 0,8 (-)0,4-2 40 1,3 0,1-2,5 1,7 (-)0,1-3,5 2,4 0,2-4,6 0,9 (-)0,7-2,5 1,2 (-)1,1-4,3 (-)0,1 (-)1,5-1,3 0,2 (-)0,8-1,2 1,9 0,1-3,7 0,9 (-)0,3-2,1 45 0,9 (-)0,3-2,1 1,5 (-)0,7-3,7 2,3 0,1-4,5 0,9 (-)0,9-2,7 1,1 (-)1,8-4 (-)0,1 (-)1,7-1,5 0,2 (-)0,8-1,2 1,8 (-)0,2-3,8 1 (-)0,2-2,2 50 1,2 (-)0,4-2,8 1,3 (-)1,2-3,8 2,2 0-4,4 1,3 (-)0,7-3,3 1,1 (-)2-4,2 0 (-)1,6-1,6 0,2 (-)0,8-1,2 1,6 (-)0,4-3,6 1 (-)0,2-2,2 p value within 0,678 0,893 0,31 0,26 0,89 0,043 0,314 0,893 0,31 the group p value within 0,77 0,61 0,57 the group 74 Table L. Rotation of the humeral head centre in the glenohumeral joint (relative motion) at Hawkins sign in patients at inclusion and after mean 38 months follow up after physiotherapy, open surgery or artroscopic surgery. Recorded values in degrees of rotation P-valeus within and between the diffrent treatmengroups. treatment options open surgery arthroscopic surgery physeotheraphy P-values rotation mean (°) CI-95% mean (°) CI-95% mean (°) CI-95% Between the groups after inclusion flexion(-), extension(+) 13 (2 - 23) -32 (-86 - 22) -7 (0- 3) follow up 0,71 flexion(-), extension(+) 14 (1 - 27) 7 (-31 - 46) -13 (-2- 3) p-value 0,735 0,345 0,953 after inclusion utåtrot.(-), inåtrotation(+) 48 (35 - 62) 64 (43- 86) 63 (53 - 74) follow up 0,19 utåtrot.(-), inåtrotation(+) 56 (42 - 71) 55 (37 - 72) 46 (34 - 58) p-value 0,237 0,893 0,86 after inclusion abd(-), adduktion(+) -54 (-62 - -45) -16 (-65 - 34) -46 (-70 - -22) follow up 0,23 abd(-), adduktion(+) -70 (-88 - -51) -58 (-68 - -48) -35 (-61 - -9) p-value 0,063 0,225 0,515 Table M. Proximal translation of the humeral head inside the glenohumeral joint (relative motion) at Hawkins sign in patients after inclusion and at mean follow up 30 months including physeotherapi, open surgery and artroscopic surgery. Values in mm. P-valus within and between groups treatment options Open surgery artroscopic surgery physeotherapi P-values translation mean (mm) CI-95% mean (mm) CI-95% mean (mm) CI-95% Between the groups after inclusion lateralt(-), medialt(+) 1.6 (-0.9 - 4.1) 1.2 (-0.8 - 3.3) 1.4 (-0.4 - 3.1) follow up 0.28 lateralt(-), medialt(+) 0.2 (-2.6 - 3.0) 3.3 (-2.3 - 9.0) 0.3 (-2.2- 2.7) p value 0.31 0.465 0.038 after inclusion distalt(-), proximalt(+) 2.3 (0.8-3.8) 1.5 (-0.3 - 3.3) 2.8 (0.3 - 5.3) follow up 0.75 distalt(-), proximalt(+) 3.6 (1.5 - 5.1) 2.5 (0.7- 4.2 3.1 (0.8 - 5.4) p-value 0.128 0.465 0.594 after inclusion posteriort(-), anteriort(+) 1,5 (-0.6 - 3.5) 2.2 (-0.2 - 4.6) 2.4 (0.1 - 4.8) follow up 0.76 posteriort(-), anteriort(+) 1.8 (0.3 - 3.3) 1.9 (-2.7 - 6.6) 1.7 (-0.3- 3.8) p-value 0.612 0.715 0.857 Table N. The clinical outcome evaluated by Constant score 75 (Constant score without strength test) and its subscore for pain(points) 3,6 12, and 24 months after 3 diffrent treatement options (physiotherapy, open surgery or artroscopic surgery). P-valus within and between the treatment groups at 24 months follow-up treatment options Open surgery artro- surgery physiotherapy Open surgery artro- surgery physiotherapy score con stant 75 con stant 75 con stant 75 pain pain pain after treatment (month) median range valid median range valid median range valid median range valid median range valid median range valid 0 43 (12-54) 7 38 (18-48) 5 47 (39-68) 7 1 (1-9) 7 2 (0-3) 5 3 (0-15) 7 3 55 (33-67) 3 60 (24-68) 5 49 (60-74) 3 6 (3-10) 7 10 (4-10) 5 6 (3-8) 3 6 56 (51-70) 5 66 (42-75) 5 43 (60-78) 5 7 (3-13) 6 10 (0-15) 5 6 (0-10) 5 12 65 (23-73) 5 75 (52-75) 4 41 (37-80) 5 11 (4-15) 7 15 (3-15) 4 8 (1-9) 5 24 65 (20-66) 5 75 (49-75) 5 54 (45-80) 5 12 (2-15) 7 15 (4-15) 5 8 (3-9) 5 P-values within the group 0,18 0,04 0,14 0,06 0,04 0,23 arhtroscopy versus physiotherpy (p-value) 0,008 open surgery versus physiotherpy (p-value) 0,048 0,43 arhtroscopy versus open surgery (p-value) 0,6 arhtro. and op surg ver physiotherpy (p-value) 0,025 75